Oxylabs
自举跑出的网页数据基础设施已有真实规模,但当前价格已经计入尚未完全公开的质量假设
Oxylabs 看起来是真正受益于 AI 时代的网络数据基础设施赢家;但当前 $3.6B 估值已经计入明显质量,公开证据留下的超额认知不足以支撑买入判断。
封面要素
公司概况
Oxylabs 是一家总部位于 Vilnius 的网页数据基础设施公司,创立于 2015,源自 Tesonet 生态。它最早是高端代理服务商,后来扩展成更完整的产品栈,覆盖住宅、移动、ISP 和数据中心代理,Web Scraper API、Web Unblocker,以及更新的 AI 导向工具。公开证据支持其已有可观商业规模——到 July 2026 约 $350M ARR、350,000+ 客户或技术团队——并且在零售情报、SEO 数据和公共部门执法中有真实客户部署。公司在 July 2026 以 $3.6B 估值从 Warburg Pincus 获得首笔外部资本。
- 成立时间
- 2015-01-01
- 创始人
- Julius Černiauskas
- 创立地点
- Vilnius, Lithuania
- 总部
- Vilnius, Lithuania
- 产品
- 一个高端网页数据平台,把代理基础设施、解封锁工具、抓取 API 和面向 AI 的抽取工具组合起来,帮助企业和开发者规模化访问并结构化公开 Web 数据。
- 客户
- 面向电商情报、AI、网络安全、SEO 及其他数据密集工作流中的企业和技术成熟买家,并通过 Webshare 覆盖更长尾的自助客户。
- 商业模式
- 基础设施和软件按用量与订阅收费,卖点是访问可靠性、合规姿态、支持,以及更高阶的网页数据工具。
- 阶段
- Private; first institutional round
- 融资情况
- July 2026 从 Warburg Pincus 获得 $130M 投资,估值约 $3.6B;自举增长十年后首次引入外部资金。
执行摘要
主要优势
- $350M ARR 和 350,000+ 客户 / 技术团队说明,Oxylabs 已经达到有分量的后期私营规模。
- 公司把自己定位成 AI 时代实时网页访问基础设施,而不是狭窄的大宗代理卖家。
- 来自 Wiser、Epicup 和 Environmental Protection Department 的公开客户证据支撑真实生产使用。
- 重合规品牌、KYC 和信任中心披露,让 Oxylabs 与声誉较弱的代理服务商拉开距离。
主要风险
- NRR、毛利率、现金烧钱和客户集中度都没有公开数据,收入质量仍披露不足。
- 平台封堵、出版方变现,以及隐私或 AI 监管,都可能迅速压缩利润率和估值倍数。
- 客户证据分化:企业用户看重性能,小客户则暴露出更多定价和支持摩擦。
- 当前估值约为 10.3x ARR,相比低倍数的上市数据供应商已经不再明显便宜。
未决问题
- 没有截至 runDate 的经审计 ARR 桥接,无法看清 $350M 数字到底包含什么。
- 没有公开披露毛利率、NRR、GRR、流失或队列留存。
- 没有公开说明头部客户集中度、分部结构或渠道经济性。
- 没有公开披露 Warburg 轮的优先权、治理权或稀释悬念。
目录
01公司概况
1.1 身份、起源与平台范围
Oxylabs 是一家总部位于立陶宛的私营网页数据基础设施公司,创立于 2015,源自 Tesonet 生态。官方和合作伙伴资料称,公司最初是高端代理服务商,之后扩展成更完整的全栈网页数据平台,覆盖住宅、移动、ISP 和数据中心代理;Web Scraper API;Web Unblocker;无头浏览器;AI Studio;Fast Search API;以及自定义数据集。官网和产品页把 Oxylabs 定位为 AI 训练、RAG、定价情报、广告验证、网络安全和旅行聚合的基础设施,而不是狭义的抓取工具。这个定位很关键,因为它把公司从有争议的代理销售商,重新定义为智能体 AI 和企业级公开数据获取的使能层。同一批官方材料也强调伦理、保险覆盖和企业支持,说明管理层希望外界按可靠性、合规和产品栈广度来评价公司,而不是只看低成本访问。[CO001, CO002, CO003, CO004, CO005, CO018]
| 指标 | 公开数值 / 状态 | 日期 / 版本时点 | 置信度 | 来源说明 |
|---|---|---|---|---|
| 成立 | 2015 | 当前叙事 | 高 | 官方公司资料与 Warburg 来源一致 |
| 总部 | 立陶宛维尔纽斯 | 当前叙事 | 高 | 官方与合作伙伴来源一致 |
| 最近一轮融资 | $130M 成长投资 | 2026-07-09 | 高 | Warburg + 公司 + 媒体 |
| 投后估值 | $3.6B | 2026-07-09 | 高 | Warburg + 公司 + 媒体 |
| ARR | $350M | 2026-07-09 | 中 | 投资报道中的公司说法 |
| 客户 | 350,000+ 客户 | 2026-07-09 | 中 | 公司 / 合作伙伴新闻稿口径;定义不清 |
| 客户 / 客户标识 | 全球 15,000+ 客户 | 当前官网首页 | 中 | 官网首页口径看起来比客户总数更窄 |
| 员工数 | 500+ 专家 | 当前合规页面 | 中 | 合规页面表述,不是 HR 申报文件 |
| 住宅 IP 池 | 175M+ IP,覆盖 195 个国家 | 当前产品页 | 高 | 官方产品页 |
| 专利组合 | 160+ 项专利 | 2026-07-09 | 中 | 公司在融资公告中声称 |
公开指标混合了全公司用户、客户和运营规模披露;ARR、客户数、员工数和专利均由管理层报告,并非独立审计。
[CO001, CO002, CO009, CO010, CO013, CO014]Oxylabs 把 Tesonet 来源的人才、代理基础设施、合规控制、AI 时代用例和 2026 年机构资本连成一个运营系统。
[CO003, CO004, CO005, CO018, CO022, CO026]1.2 管理层交接、治理与关键人依赖
2026,Oxylabs 经历了一次重要管理层交接。公司宣布,创始人兼长期 CEO Julius Černiauskas 将转任董事长,Vytautas Savickas 将出任 CEO,负责日常运营。这个安排看起来是在公司接受首笔外部资本的同一时间,把创始人层面的战略影响力和运营执行拆开。不过,公开治理披露仍不完整:Oxylabs 披露了 CFO,以及风险、法务和治理负责人,但在已审阅来源中没有公布投资后完整董事会名单、Warburg 董事席位权利或详细股东控制信息。因此,关键人和治理尽调存在中等风险。Julius 作为创始人兼董事长仍是战略核心;Savickas 现在是公司 AI 时代投资逻辑的外部代表;CFO Jurgis Rudgalvis 在收购和规模化战略上被频繁引用。公开记录能支持领导层梯队已经拓宽,但不足以证明新资本结构背后的具体治理机制。[CO006, CO007, CO008, CO026, CO027, CO040]
| 人物 | 公开来源中的当前职务 | 职责证据 | 治理相关性 | 关键人 / 尽调说明 |
|---|---|---|---|---|
| Julius Černiauskas | 创始人;正从 CEO 转任董事会主席 | 领导层交接公告 | 创始人影响力与战略连续性 | 需获取完整董事会章程和持续控制权 |
| Vytautas Savickas | 首席执行官(CEO) | 领导层交接公告;Warburg 引述 | 日常运营负责人和主要对外发言人 | 融资后扩张的关键执行负责人 |
| Jurgis Gabrielius Rudgalvis | 首席财务官(CFO) | Warburg 与 Tesonet 交易报道 | 资本配置、并购、投资人接口 | 需确认 CFO 任期、汇报节奏和融资权限 |
| Denas Grybauskas | 首席治理与战略官 | 风险 / 合规页面 | 显示治理职能正在制度化 | 需确认与董事长和法务团队的职责重叠 |
| Vaidotas Šedys | 首席风险官 | 风险 / 合规页面 | 监督 KYC、控制和运营风险 | 对合规护城河重要,但外部细节很少 |
本表只覆盖公司和合作伙伴披露中公开具名的领导者;完整董事会名单、独立董事和 Warburg 观察员权利尚未公开披露。
[CO006, CO007, CO008, CO026, CO040]1.3 融资历史、资本结构与战略股东
资本结构上最关键的事实是,Oxylabs 从 2015 一直自举到 July 9, 2026;当天 Warburg Pincus 通过 Capital Solutions Founders Fund 以约 $3.6B 估值投资 $130 million。多家官方和独立来源都把这一轮称为公司的首笔外部投资,也是立陶宛创业生态的标志性事件。这轮融资也正式把公司从创始人自有资金驱动的规模化公司,推向外部资本支持的增长平台。公开披露显示,Goldman Sachs 是 Oxylabs 的独家配售代理,Deloitte Lithuania 担任财务和税务顾问,多家国际律所为交易双方提供法律服务。管理层也明确把新资本和未来收购、合作伙伴关系挂钩,延续此前 2022 收购 Webshare、2025 收购 ScrapingBee 的路径。由于所有权比例、清算优先权或 Warburg 治理权利仍没有公开拆解,投资者可以把标示估值视为已验证,但详细控制图仍不完整。[CO009, CO010, CO011, CO012, CO023, CO025]
| 利益相关方 | 角色 / 类型 | 重要性 | 公开可见控制权 / 经济权益 | 尽调问题 |
|---|---|---|---|---|
| Warburg Pincus | 成长投资人 | 提供首笔外部资本和外部背书 | 通过 WPCS Founders Fund 按 $3.6B 估值投资 $130M | 获取持股比例、董事会权利、优先权和信息权 |
| Tesonet 生态 | 创始生态 / 创业孵化平台 | 起源网络、战略人才池和信誉来源 | 所审来源未披露直接经济权益 | 确认当前所有权和关联方安排 |
| Julius Černiauskas | 创始人兼董事长 | 战略连续性和潜在股东影响力 | 已披露董事会主席角色;经济持股未披露 | 确认投票控制权和创始人锁定安排 |
| Vytautas Savickas | 运营 CEO | 2026 年交接后的主要执行负责人 | 未披露股权所有权数据 | 确认激励和继任设计 |
| Webshare | 2022 年收购标的 | 扩大自助服务和低价产品覆盖 | 收购已完成;Webshare 博客称价格不变 | 复核整合经济性和交叉销售表现 |
| ScrapingBee | 2025 年收购标的 | 扩大面向开发者的抓取 API 组合 | 被提及为战略性产品组合收购 | 复核收购价、留存和整合影响 |
公开来源能确认参与方,但没有披露股权结构比例、优先股堆叠或交割后治理权利。
[CO009, CO011, CO012, CO023, CO025, CO037]1.4 规模指标、合规姿态与公开经营足迹
Oxylabs 的公开规模指标亮眼,但需要谨慎解读。公司的投资公告和 Tesonet 报道称,Oxylabs 已达到 $350 million ARR,客户超过 350,000;官网则称全球 15,000+ 客户信任它,合规页面称公司有 500+ 专家。最可能的解释是,这些数字描述的群体不同——比如开发者或技术团队总用户数,相对于付费企业客户——但公司没有在已审阅公开来源中给出清晰定义。无歧义的是平台本身的规模:175M+ 住宅 IP、20M+ 移动 IP、1.9M+ 专用数据中心 IP 池,以及产品页展示的企业级定价、正常运行时间和定向能力主张。Oxylabs 也通过 ISO 27001、SOC 2 引用、信任中心、法务 / 合规人员配置、KYC 要求和伦理叙事,持续强化合规与信任,用来区别于声誉较低的代理服务商。[CO013, CO014, CO015, CO016, CO017, CO019]
| 领域 | 公开披露规模 / 功能 | 核心买方需求 | 证据 | 含义 |
|---|---|---|---|---|
| 住宅代理 | 175M+ IP、195 个国家、0.41s 响应,$6/GB 起 | 高难目标抓取、广告验证、AI 数据采集 | 官方产品页 | 核心规模护城河和高端定位 |
| 移动代理 | 20M+ IP、140+ 个国家、1.1s 响应 | 移动端专属、地理敏感、反机器人压力高的任务 | 官方产品页 | 支撑高摩擦目标和本地化数据 |
| 专用数据中心代理 | 1.9M+ 池、188 个国家、99.9% 成功率 / 正常运行时间 | 高性价比大规模抓取 | 官方数据中心页面 | 为技术栈加入低成本吞吐层 |
| Web Scraper API | $49/month 起,提供结构化输出和 AI 集成 | 无需运营代理栈即可托管采集数据 | 官方 API 页面 | 把价值上移到交钥匙 API |
| 信任 / 合规 | ISO 27001、SOC 2 引用、KYC、信任中心、伦理准则 | 企业采购和声誉防御 | 风险 / 合规和信任来源 | 合规定位是进入市场护城河的一部分 |
| 客户证明 | 跨平台 1,150+ 条正面评价和案例研究 | 为企业和自助买方提供社会证明 | 产品页、评价、案例研究 | 支撑高端定价,但不足以消除支持投诉 |
各行汇总了当前 Oxylabs 页面中的公开产品和信任披露;除非明确说明,数值均为公司运营口径,不是第三方审计。
[CO018, CO019, CO020, CO021, CO022, CO026]| 指标族 | 报告数值 | 来源 | 可能的范围线索 | 尽调含义 |
|---|---|---|---|---|
| ARR | $350M | 融资公告 / Tesonet / 媒体 | 管理层报告的经常性收入水平 | 需要数据室验证和收入质量拆解 |
| 客户 | 350,000+ 客户 | Warburg / Tesonet / TNW | 可能包含广义用户或开发者账户基数 | 明确付费、活跃与注册账户口径 |
| 客户 | 全球 15,000+ 客户 | 官网首页和合规页面 | 可能是更窄的付费 / 标识客户数 | 做估值前,先与 350k 客户数对齐 |
| 员工数 | 500+ 专家 | 合规页面 | 可能是员工或专家团队口径 | 与 HRIS、LinkedIn 和区域实体交叉核验 |
| 专利 | 2025 年 150 项,2026 年 160+ 项 | 影响报告 vs 2026 年 7 月融资帖子 | 连续增长看起来可信 | 需要专利质量和司法辖区图谱,不能只看数量 |
本表有意对比一组相关公开指标;在所审来源中,这些指标的定义并未完全解释。
[CO013, CO014, CO015, CO016, CO017, CO023]1.5 里程碑、AI 定位与反向信号
公开时间线显示,Oxylabs 一直从代理基础设施稳步拓展到面向 AI 的网页数据系统。关键里程碑包括创立于 2015、2022 收购 Webshare、2025 收购 ScrapingBee 并推出 AI Studio、March 2026 领导层重组,以及 July 2026 Warburg 融资。公司也通过 AI Predictions 2026、DEVWorld 和 AI Engineer World’s Fair 明显向开发者和 AI 构建者营销,这支持了 Oxylabs 把智能体 AI 视为主要需求驱动的说法。反向记录没有到生死攸关的程度,但仍相关:Trustpilot、G2 和 Capterra 都出现了围绕定价、支持响应、退款、KYC 摩擦和小用户可负担性的批评;TNW 也明确指出,网页数据业务仍带有法律和声誉包袱。这些信号不会推翻增长故事,但提醒投资者,Oxylabs 的高端估值取决于它能否在扩张一个持续受到客户、网站和监管机构审视的行业时,守住合规可信度。[CO023, CO025, CO028, CO029, CO030, CO031]
| 日期 | 事件 | 类型 | 金额 / 状态 | 参与方 | 含义 |
|---|---|---|---|---|---|
| 2015 | Oxylabs 在 Tesonet 生态内于维尔纽斯成立 | 创立 | 公司成立 | Tesonet 生态;Julius Černiauskas | 自举叙事的起点 |
| 2022-09 | 完成收购 Webshare Software | 合作 | 收购完成 | Oxylabs;Webshare | 拓展低价自助服务覆盖 |
| 2025 | 收购 ScrapingBee;推出 AI Studio | 产品 | 产品组合扩张 | Oxylabs;ScrapingBee | 拓宽开发者工具和面向 AI 的产品集 |
| 2025 | 专利组合达到 150 项 | 规模 | 里程碑披露 | Oxylabs | 显示 IP 持续积累 |
| 2026-03 | 创始人 Julius Černiauskas 转任董事长;Vytautas Savickas 出任 CEO | 治理 | 领导层重组 | Oxylabs 领导层 | 将战略与运营领导拆分 |
| 2026-07-09 | Warburg Pincus 通过 WPCS FF 投资 $130M | 融资 | $130M | Warburg Pincus;Oxylabs | 首笔外部资本和机构背书 |
| 2026-07-09 | 估值达到约 $3.6B | 规模 | $3.6B | Warburg;Oxylabs | 确认独角兽地位和国家层面的重要性 |
| 2026-07 | 公司称 ARR 为 $350M,客户数 350,000+ | 规模 | 当前规模指标 | Oxylabs;Tesonet;媒体 | 成为后续财务和估值工作的基线 |
| 2026-07 | AI 事件和 AI 预测传播扩张 | 产品 | 进入市场 / 思想领导力 | Oxylabs | 显示正主动转向智能体 Web 叙事 |
里程碑仅限已抓取来源中可直接观察到的事件;若来源只描述月份或年份,部分日期仍为近似。
[CO001, CO003, CO006, CO007, CO009, CO010]公司公开时间线显示:自举增长、两次收购、2026 年领导层交接,以及 2026 年 7 月融资事件共同坐实其独角兽身份。
[CO001, CO003, CO006, CO007, CO009, CO010]1.6 要点
02市场分析
2.1 市场边界、纳入支出与替代方案
Oxylabs 所在市场比泛泛的“代理服务器”更宽,但比所有企业数据基础设施更窄。它包括采集公开 Web 数据的软件和服务、住宅和数据中心代理访问、托管抓取 API、反机器人绕过、浏览器自动化,以及越来越多的开放 Web 和 AI 事实锚定基础设施。TBRC 的分层覆盖软件与服务、多种部署模型,以及价格监测、市场研究、数据聚合和竞争情报等应用。OpenWebSearch 和 Cloudflare 又加了一层边界:开放 Web 索引和程序化内容访问,正在成为 AI 系统相邻的基础设施选择。主要替代方案包括官方 API、授权数据集、人工研究、自建爬虫栈,以及开放索引或搜索供应商访问。因此,Oxylabs 并不只是和其他代理供应商竞争;它也在和内部工程自建、平台级替代方案竞争,后者会在灵活性与合法性、新鲜度或成本可预测性之间取舍。[CM001, CM005, CM006, CM007, CM008, CM026]
| 细分 / 类别 | 纳入支出 | 排除支出 | 买方 / 付款方 | 对 Oxylabs 的重要性 |
|---|---|---|---|---|
| 网页抓取软件 | 抓取工具、抽取平台、解析层、托管抓取 API | 通用分析工具、非 Web ETL、与抓取无关的广告技术 | 数据团队、产品团队、增长团队 | Oxylabs API 和工具最接近的狭义类别 |
| 代理基础设施 | 住宅、数据中心、ISP 和移动代理访问 | 消费者 VPN 和无关隐私工具 | 工程、反欺诈、SEO、电商、威胁情报买方 | 构成 Oxylabs 直接变现的连接层 |
| 反机器人绕过 / 访问层 | 解封器、无头浏览器访问、会话和身份管理 | 只卖给防守方的纯网站安全产品 | 高级数据采集团队和 AI 运营方 | 对高难目标采集和高端定价至关重要 |
| 开放 Web 索引 / 搜索基础设施 | 索引访问、开放 Web 数据库、AI / 搜索爬取框架 | 传统搜索广告和消费者搜索 UX | 研究人员、AI 构建者、搜索和 RAG 团队 | 创造与直接爬取相邻的替代品和补充品 |
| 授权或官方数据访问 | 官方 API、结构化数据馈送、授权数据集 | 未经授权的私有数据、封闭企业数据库 | 采购主导的企业、受监管买方 | 合规或可靠性压过原始灵活性时,这是主要替代项 |
| 自建采集栈 | 内部爬虫工程、数据 QA、反机器人运营、浏览器自动化 | 现成 SaaS 支出 | 大型平台和成熟数据团队 | 说明技术深度和 TCO 为何影响市场选择 |
这个市场边界有意纳入买方可能与 Oxylabs 一起评估的相邻基础设施;但并不把所有数据基础设施都视为同一个支出池。
[CM001, CM005, CM006, CM007, CM028, CM047]| 替代 / 相邻选项 | 买方得到什么 | 相较完整抓取栈失去什么 | 典型买方 | 为什么 Oxylabs 仍重要 |
|---|---|---|---|---|
| 官方 API | 合法性高、数据模式稳定、有支持 | 覆盖有限、配额限制、字段缺失、范围扩张更慢 | 受监管企业和产品团队 | 许多重要目标没有暴露足够 API |
| 授权数据集 | 数据权利和刷新条款可预测 | 灵活性低,自定义覆盖更弱 | 采购主导的企业、金融、研究 | 自定义目标覆盖和及时性仍需要直接采集 |
| 人工研究 | 工具支出低,有上下文判断 | 无法规模化、新鲜度弱、人工成本高 | 小团队和一次性项目 | 刷新频率一上来,自动化就胜出 |
| 自建爬虫栈 | 控制力最大,逻辑可定制 | 工程负担重,反机器人维护持续消耗 | 大型平台和数据原生公司 | 高端提供商缩短见效时间,降低运营风险 |
| 开放 Web 索引访问 | 搜索 / R&D 用例的广覆盖语料 | 研究许可或索引约束,按目标定制的控制力较弱 | 研究人员、主权搜索团队、AI 实验室 | 精准、交易型或实时目标上,Oxylabs 仍更合适 |
| 低价商品化代理 | 便宜访问和简单自助 | 合规弱、IP 质量不稳定、支持深度较低 | 自由职业者和业余自动化玩家 | 企业买方往往需要可靠性、KYC 和托管工具 |
表中区分两类选择:一类替代项会消除抓取需求,另一类只是把实现层挪走。
[CM001, CM018, CM028, CM044, CM045, CM046]品类变现按顺序展开:先是开放网络内容,最后变成可用的企业或 AI 工作流,每一层都有阻碍。
这是一张流程图,不是按时间缩放的漏斗;它突出公开来源中可见的价值链和摩擦点。
[CM001, CM022, CM029, CM030, CM031, CM038]2.2 市场规模口径与区域结构
公开市场规模估计指向同一个方向——快速增长——但绝对数差异很大。The Business Research Company 估算 2026 的网页抓取市场为 $1.17 billion,并预测到 2030 达到 $2.28 billion;Mordor Intelligence 则把 2026 市场估为 $1.56 billion、2031 市场估为 $3.49 billion。这些差异大到不能忽视,可能反映服务、基础设施和相邻产品的纳入规则不同。ScrapingDog 和 WorldMetrics 在纳入另类数据、主流分析采用或更宽软件类别时,引用的背景数字更大。实际结论是,投资者不应锚定单一 TAM 幻灯片;更合理的做法是把市场看成分层外壳,其中狭义网页抓取软件是低个位数十亿美元级别品类,但更广义的公开 Web 数据栈更大,而且因为服务 AI、定价、威胁情报和搜索应用而越来越具战略性。TBRC 还称北美是 2025 的最大区域,亚太是增长最快区域,这一格局与西方成熟企业需求、亚洲数字商业和平台规模扩张相一致。[CM002, CM003, CM004, CM009, CM017, CM042]
| 视角 | 来源 / 年份 | 地理范围 | 数值 | 方法信号 | 置信度 / 局限 |
|---|---|---|---|---|---|
| 全球网页抓取市场 | TBRC 2026 | 全球 | 2026 年 $1.17B;2030 年达 $2.28B | 软件 + 服务市场报告,按应用和垂直行业细分 | 中;软件 / 服务边界较窄 |
| 全球网页抓取市场 | Mordor Intelligence 2026 | 全球 | 2026 年 $1.56B;2031 年达 $3.49B | 行业报告,列出主要供应商和 CAGR | 中;供应商边界比 TBRC 更宽 |
| 更宽的行业语境 | ScrapingDog 2026 | 全球 | 多个研究来源引用 $2.2B–$3.5B 的市场轨迹 | 综合多份研究报告和另类数据语境 | 中低;二手汇总,不是原始模型 |
| 另类数据语境 | ScrapingDog 2026 | 全球 | $4.9B 另类数据语境,年增长 +28% | 与抓取驱动数据集相邻的市场口径 | 中低;相邻口径,不是直接抓取支出 |
| 代理市场增长视角 | Cybernews / Proxyway 2026 | 全球 | 部分主要提供商同比增长 50%+ | 参与者问卷加基准评测评论 | 中;供应商自报增长,不是经审计市场规模 |
| 开放 Web 索引规模 | OpenWebSearch 2026 | 欧洲 / 研究用途 | 9.14B 个 URL、1 TiB/day、1 PB 数据存储 | 开放索引替代方案的运营规模,不是收入总可用市场(TAM) | 中;规模视角,不是支出视角 |
这些视角有意采用不同市场边界;只能方向性比较,不能相加。
[CM002, CM003, CM004, CM009, CM017, CM020]从狭义抓取软件支出扩展到更广的公开网络数据和 AI 基础设施范畴,可投资市场随之变宽。
各层刻意采用不同市场口径,而不是一套可相加的层级;目的在于展示边界扩张,而非加总市场规模。
[CM002, CM003, CM017, CM028]公开 2026 年价格区间显示,住宅代理市场分层明显,高端性能和价值定价并存。
所有行都用 USD per GB 表示住宅代理定价;区间结合了 2026 年基准 / 对比页面中的公开入门价和折扣后大用量价格点。
[CM042, CM043, CM045, CM046]2.3 买方、用户与付款方分层
这个市场由多个群体采购,他们关心的东西并不相同。大型企业和 SMBs 都出现在 TBRC 的分层中;Apify 的 2026 调查则显示,35.8% 受访者是自由职业者,49.1% 在初创公司或 SMBs 工作。实际采购群体大致围绕电商和定价情报团队、增长和线索生成团队、市场研究和另类数据团队、网络安全和欺诈团队,以及需要新鲜 Web 数据来训练、检索或驱动智能体的 AI/LLM 产品团队。预算负责人也随之变化:产品或工程负责人可能负责抓取 API,营销负责人可能负责线索生成抓取,欺诈或威胁情报负责人可能负责对反机器人敏感的数据采集。这种多样性很重要,因为 Oxylabs 在部分细分市场可以靠企业合规和高端性能取胜,而低成本自助供应商会在自由职业者或小型自动化团队中获胜。销售周期和付费意愿也会随着工作流关键性和法律敏感度显著变化。[CM006, CM007, CM008, CM012, CM018, CM019]
| 细分领域 | 主要用户 | 预算所有者 / 付款方 | 工作流 | 采用触发因素 |
|---|---|---|---|---|
| 电商与零售情报 | 定价或品类分析师 | 电商运营 / 选品负责人 | 价格监控、品类跟踪、卖家情报 | 需要快速、本地化的竞品数据 |
| 增长 / 线索生成团队 | 增长运营或 RevOps 分析师 | 市场 / 增长负责人 | 线索补全、潜客开发、市场地图绘制 | 需要自动化采集公开来源的重复工作 |
| 另类数据 / 市场研究 | 分析师和数据科学家 | 研究负责人或业务单元总经理 | 为市场模型和仪表盘做大规模采集 | 需要广度、新鲜度和可重复性 |
| 网络安全 / 欺诈 / 信任 | 威胁情报或欺诈分析师 | 安全、信任或风险预算所有者 | 威胁监控、暴露测试、滥用调查 | 需要稳定触达敌对或持续变化的目标 |
| AI / LLM / RAG 产品团队 | ML 工程师和平台团队 | 产品 / 工程 / AI 平台预算 | 训练、事实锚定、实时检索、智能体工作流 | 需要带治理的结构化实时数据 |
| 自由职业者和 SMB 自动化团队 | 独立开发者或小型运营团队 | 创始人或个人买家 | 轻量抓取或代理任务 | 需要低启动成本和自助工具 |
买方群体综合自 TBRC 的垂直 / 应用细分,以及 Apify 调研中自由职业者、初创公司、SMB 和成熟抓取从业者的构成。
[CM006, CM007, CM008, CM012, CM018, CM019]同一宽泛买方群体,一旦考虑合规敏感度和工作流关键性,就会拆成截然不同的决策权和付费意愿画像。
[CM006, CM007, CM008, CM012, CM018, CM019]2.4 增长驱动与采用顺风
多个来源都指向持久的需求扩张。TBRC 把数字化转型、AI 和机器学习训练数据集、云平台、实时趋势分析列为主要增长驱动。Apify 调查显示代理和基础设施支出在上升;Cybernews/Proxyway 称,AI 公司追求更大容量和更高频数据刷新,一些主要供应商同比增速超过 50%。OpenWebSearch 和 Cloudflare 从不同角度强化了同一个宏观判断:AI 应用越来越需要实时访问 Web,而不是陈旧索引,无论是通过欧洲开放索引,还是直接爬取并付费的机制。Oxylabs 自己的 AI Predictions 材料也与此一致,认为 AI 智能体正在成为工作流操作者,数据来源、治理和实时访问比单纯模型新颖性更重要。能把代理规模、API 抽象、合规和企业支持组合起来的供应商,会获得有利的需求背景。[CM010, CM011, CM012, CM013, CM014, CM015]
| 驱动因素 / 约束 | 方向 | 时点 | 证据 | 影响 |
|---|---|---|---|---|
| AI 训练和智能体检索需求 | 正向 | 当前 | TBRC, Oxylabs, Cloudflare, OpenWebSearch | 扩大对实时、结构化公共 Web 数据的需求 |
| 数字化转型和自动化决策 | 正向 | 当前 | TBRC 市场驱动因素章节 | 让品类价值不止于 AI 热潮 |
| 代理和基础设施支出上升 | 正向 | 当前 | Apify 2026 调研 | 表明买方愿意为规模化可靠性付费 |
| 开放 Web 和数字主权倡议 | 正向 | 新兴 | OpenWebSearch / 开放网络索引 | 创造新应用,并让公共 Web 数据访问更具正当性 |
| 出版方将 AI 爬取访问变现 | 负向 | 新兴 | Cloudflare 按爬取付费 | 可能抬高输入成本,压缩免费可爬取范围 |
| 反机器人技术升级和分布式攻击 | 负向 | 当前 | DataDome 2026 研究 | 推动买方转向高端供应商,或导致项目失败 |
| 监管可追溯性和数据集义务 | 负向 | 当前至 2026 年 8 月 | AI Act 摘要和法律文本 | 利好合规供应商,但抬高上手摩擦 |
| 代理层商品化和低进入门槛 | 混合 | 当前 | Cybernews/Proxyway 和 AIMultiple 定价 | 压缩商品化利润率,同时高端层扩张 |
同一股力量可能帮到一层、伤到另一层;例如监管可以提升品类正当性,也会增加采集摩擦。
[CM010, CM011, CM012, CM013, CM020, CM021]2.5 监管、发布方与反机器人约束
同一个市场也在变难。DataDome 把抓取描述为数据所有者面临的数十亿美元级问题,并展示攻击者如何使用住宅代理、伪造用户代理和分布式攻击。其 2026 的智能体身份研究认为,大多数 AI 智能体仍不能正确表明身份,大多数网站也无法区分合法智能体和伪造智能体。Cloudflare 推出按爬取付费,是条件收紧的另一个信号:发布方越来越希望默认阻止 AI 爬虫或向其收费,而不是允许免费访问。AI Act 又加了一层合规要求,对高风险和 GPAI 相关系统施加数据集质量、可追溯性和透明度义务,同时明确禁止为人脸识别数据库无差别抓取互联网数据。合在一起,这些力量意味着品类未来的利润池可能从无差异代理转售,转向能提供合法来源、智能体认证、自适应反机器人处理和更清晰企业治理的供应商。Oxylabs 的高端定位符合这个方向,但同一趋势也会抬高客户上手摩擦和法律审查的重要性。[CM022, CM023, CM024, CM025, CM029, CM030]
2.6 要点
03竞争格局
3.1 格局:直接同业、邻近玩家、替代方案与潜在进入者
Oxylabs 周围的竞争集,比简单的 Bright Data 对 Oxylabs 更宽。直接高端同业包括 Bright Data;在买家需要大规模住宅池加托管解封锁能力时,SOAX 和 Decodo 也在一定程度上进入同一集合。但邻近集合同样重要:Apify 卖工作流自动化和 AI 工具市场,ScraperAPI 和 ScrapingBee 卖托管抓取结果而不是原始流量,Webshare 或 IPRoyal 等低成本代理供应商面向不需要高端企业支持的买家。Proxyway、AIMultiple、PrivateProxyGuide、BestProxyCompare 和 Distill 的独立排名都强化了同一个市场结构:Oxylabs 同时面对高端全栈竞争者、更便宜的自助供应商、API 优先替代方案,以及内部自建或维持现状的选择。这个更宽的格局很重要,因为它说明主要投资判断不是 Oxylabs 能否在基准速度上击败某一个竞争对手,而是当买家在成本、合规、托管工作流便利性和原始基础设施控制之间取舍时,它能否保持差异化。[CP001, CP003, CP012, CP014, CP015, CP020]
| 类别 | 代表供应商 | 核心任务 | 典型买家 | 对 Oxylabs 的含义 |
|---|---|---|---|---|
| 直接高端同业 | Bright Data、SOAX | 带企业支持的高性能公网数据访问 | 大型数据团队和采购主导型企业 | 这是最接近的正面对比组,用来比较利润率和功能 |
| 准高端性价比挑战者 | Decodo | 接近高端的代理功能,配更容易自助采购的价格体系 | SMB、代理商、技术团队、中端市场买家 | 当 Oxylabs 被认为太贵或销售太重时,可以赢走交易 |
| 低价自助供应商 | Webshare、IPRoyal、Rayobyte | 便宜流量、测试、窄场景访问 | 个人、试点项目、价格敏感买家 | 压低低端市场,限制 Oxylabs 从简单用例变现 |
| 托管抓取 API | ScraperAPI、ScrapingBee | 无需直接管理代理即可获取结构化数据 | 想快速落地的开发者 | 把竞争从网络广度推向工作流简单度 |
| 自动化 / 市场平台邻近玩家 | Apify | 用于可重复数据工作流的 Actors、工具和编排 | 自动化团队、AI 构建者、开发者占比较高的组织 | 攻击的是工作流所有权,而不只是传输层 |
| 现状 / 内部自建替代 | 内部爬虫栈、混合低价代理、官方 API | 对特定目标有直接控制,或成本更低 | 大型工程组织或受监管团队 | 如果 Oxylabs 无法证明更好的 TCO 或更低运营风险,定价权会被封顶 |
版图有意纳入替代品和内部自建路径,因为买家并不只在代理供应商之间选择。
[CP001, CP015, CP019, CP020, CP032, CP036]Oxylabs 与 Bright Data 同处企业广度高、价格可负担性较低的象限;Decodo、Apify、ScraperAPI 和 Webshare 则靠更低采用摩擦或工作流便利性取胜。
轴向分数是基于已保留的定价、打包和信任证据给出的分析性序数,不代表观察到的市场份额。
[CP003, CP013, CP015, CP017, CP020, CP023]3.2 竞品画像、规模信号与战略方向
公开来源一致把 Oxylabs 放在高端梯队,而不是预算层。Oxylabs 自己的官网仍围绕 15,000+ 客户基础和 AI 时代数据工作流展开;Bright Data 则营销最大的住宅网络、最宽的相邻数据访问工具集,以及适合大型采购驱动组织的合规覆盖。Decodo 是最显眼的价值型挑战者:它把仍然很大的伦理来源代理池、自助包装、按用量付费选项和第三方“best value”认可组合起来。SOAX 介于高端和值之间,强调路由质量和可预测定价。Apify 的战略不同,因为它把代理与 AI/数据工具市场混合在一起,竞争的是工作流所有权,而不只是代理流量。Webshare、IPRoyal 和 Rayobyte 分别攻击更窄的任务——便宜入门流量、按用量付费住宅使用,或低成本数据中心和 ISP 经济性——但都没有在全栈企业广度上追平 Oxylabs。综合画像证据看,Oxylabs 属于高要求企业工作负载的短名单,但它周围的竞争者可以把部分任务拆开,以更低成本或更便利方式提供。[CP002, CP004, CP005, CP006, CP010, CP013]
| 竞争对手 | 类别 | 规模 / 融资 | 目标细分 | 差异化 | 局限 |
|---|---|---|---|---|---|
| Oxylabs | 高端代理和 Web 数据平台 | 全球 15,000+ 客户;2026 年对比中引用 175M+ 住宅 IP、2M+ 数据中心 IP | 大型企业、AI / 数据团队、对合规敏感的买方 | 企业支持、高端性能、Web Unblocker、Web Scraper APIs、AI Studio | 公开证据仍显示,自助广度和数据 / 工具市场深度弱于 Bright Data |
| Bright Data | 高端全栈 Web 数据平台 | 400M+ 住宅 IP;20,000+ 组织;认证和数据集覆盖广 | 大型企业、代理机构、AI / 数据运营方、采购主导团队 | 最大的公开网络、数据集市场、浏览器和智能体工具、广泛的合规营销 | 在较窄用例中可能显得昂贵、复杂或火力过剩 |
| Decodo | 价值型高端代理和抓取平台 | 195+ 地点的 115M+ 住宅 IP;99.92% 成功率声明;按量付费和订阅 | 开发者、SMB、中端市场抓取方、价格敏感团队 | 自助 UX 强、性价比好、最佳价值口碑 | 企业广度或数据集 / 工作流层的公开证据较少 |
| SOAX | 中端市场高端代理网络 | 第三方对比引用 155M+ IP;官方定价强调经济性可预测 | 需要细粒度地理定位和稳定路由的团队 | 过滤灵活、定价可预测、稳定 IP 定位 | 更广产品面的公开证据比 Oxylabs 或 Bright Data 更薄 |
| Apify | 自动化和抓取工作流平台 | 工具市场首页显示 53,640 个 AI 工具;2026 年对比内容仍活跃 | 开发者、自动化团队、AI 智能体、工作流构建者 | 工具市场 / Actor 生态和工作流掌控力,而不只是原始代理 | 想直接控制高端代理基础设施的买方,不一定首选它 |
| Webshare | 低价自助式代理提供商 | 永久 10 个免费代理层;低价数据中心和住宅代理定价 | 个人、试点、业余玩家、预算敏感团队 | 进入门槛最低,并有免费层 | 企业支持和高端工作流定位较弱 |
| ScraperAPI | 托管抓取 API 替代方案 | 每月免费 1,000 credits,加高端代理池和基于用量的扩展 | 想要 API 级结果、不想管代理的开发者 | 集成简单、托管重试、代理抽象 | 买方若想控制原始网络或定制基础设施设计,用处较小 |
| IPRoyal / Rayobyte | 低价住宅或数据中心替代方案 | 按量付费住宅代理(IPRoyal)和低价数据中心 / ISP 经济性,加免费试用(Rayobyte) | SMB、效果营销人员、窄用例买方 | 价格激进,上手简单 | 不是全栈高端 Web 数据平台定位 |
规模 / 融资字段混合了公开网络规模、客户数或明确披露的商业规模信号,而不是强行套用同一个指标。表中聚焦买方最可能与 Oxylabs 同时列入候选名单的公司,而不是市场上每个代理转售商。
[CP002, CP004, CP005, CP006, CP010, CP013]3.3 能力广度、定价架构与信任姿态
能力广度仍是市场顶部和商品化代理转售拉开差距的地方。Bright Data 自己的对比页和 Apify June 2026 横评都显示,Bright Data 和 Oxylabs 覆盖核心代理栈、抓取 API 和数据集,但 Bright Data 又增加了更宽的数据集市场、Web IDE、浏览器自动化和智能体工具,以及更广泛的按用量付费包装。Oxylabs 用 AI Studio、Web Scraper APIs 和 Web Unblocker 缩小了差距,但公开来源仍把它描绘得略偏销售驱动、企业形态更重。Decodo 和 SOAX 弥合了大量基础代理功能差距,同时在可访问性或价格上压低门槛。信任姿态也重要:Bright Data 明确展示 ISO 和 SOC 控制,Oxylabs 与 Decodo 则强调伦理来源和合规话术。这让定价比较很容易失真。买家比较的不是每 GB 多少美元;他们比较的是企业支持、解封锁成功率、地理定向深度、市场广度、采购友好度,以及自己想把多少抓取复杂度留在内部。[CP006, CP007, CP008, CP009, CP010, CP011]
| 购买标准 | Oxylabs | Bright Data | Decodo | SOAX | Apify | Webshare | ScraperAPI |
|---|---|---|---|---|---|---|---|
| 住宅 + 数据中心 + ISP 覆盖广度 | 强 | 很强 | 强 | 中 | 弱 | 中 | 弱 |
| 托管解封 / 反机器人层 | 强 | 很强 | 中 | unknown | 中 | 弱 | 强 |
| 数据集 / 工作流市场 | 中 | 很强 | 弱 | 弱 | 很强 | 弱 | 弱 |
| AI / 智能体工具叙事 | 强 | 很强 | 中 | 弱 | 强 | 弱 | 中 |
| 企业合规信号 | 强 | 很强 | 中 | unknown | 中 | 弱 | 中 |
| 自助性价比 / 低摩擦进入 | 中 | 中 | 强 | 中 | 强 | 很强 | 强 |
| 支持想控制原始代理的买方 | 强 | 强 | 强 | 中 | 中 | 中 | 弱 |
评分是有证据支撑的等级判断,来自保留的官方产品页和 2026 年对比文章;未知表示所审阅来源集不足以支撑有把握的评估。
[CP007, CP008, CP009, CP011, CP016, CP020]| 供应商 | 入门信号 | 单位 / 合同模型 | 包含能力 | 折扣 / 未知项 | 含义 |
|---|---|---|---|---|---|
| Oxylabs | 住宅代理价格处于高端区间;Bright Data 2026 指南显示 Web Unblocker 起价约 $9.4/GB | 销售驱动的企业打包与产品级套餐并行 | 高端代理、Web Unblocker、抓取 API、AI Studio | 批量折扣和大额用量的具体条款未完全公开 | 适合企业;对爱好者或试点买家显然没有那么优化 |
| Bright Data | 对比内容显示住宅代理约从 $2.50/GB 起;广泛支持现用现付(PAYG) | 按量计费,加产品专属套餐和试用 | 代理栈、数据集、解锁器、浏览器 / 智能体工具 | 打包范围很广,也可能显得复杂 | 全栈菜单最灵活,但复杂度会加重比价疲劳 |
| Decodo | 入门套餐住宅代理约 $3.75/GB 起;支持按用量付费 | 月度订阅加 PAYG | 住宅 / 移动 / 数据中心代理和站点解锁器 | 用量越高折扣越好;提供 3 天试用 | 最适合想要接近高端能力、但不想承担偏企业定价的买家 |
| SOAX | 官方定价强调可预测的成本;公开来源显示住宅代理价格介于高端到中端之间 | 按国家层级 / 套餐进行按量定价 | 住宅、移动、ISP 以及地理定位深度 | 保留下来的公开定价细节少于 Decodo 或 Bright Data | 主打稳定性和定位能力,而不是单纯最低价 |
| Webshare | 免费层加低成本代理定价 | 按代理和按 GB 的入门套餐 | 数据中心、静态住宅和成本更低的轮换选项 | 付费套餐广度窄于高端同业 | 适合测试、个人项目,以及不需要高端支持的买家 |
| IPRoyal | 强调住宅代理按量付费 | 现用现付(PAYG)加套餐变体 | 聚焦住宅和静态住宅 | 保留证据中企业打包可见度较低 | 对优先看重灵活性、而非完整平台广度的买家,是有用基准 |
| Rayobyte | 便宜的数据中心和 ISP 定价,加免费试用 | 按产品采用按 IP 或按 GB 模式 | 数据中心、ISP、住宅和解锁器产品 | 并非所有产品都能在工作流广度上匹配 Oxylabs | 显示低价基础设施如何压制商品化层 |
| ScraperAPI | 免费额度 1,000 credits/月;文档和网页引用付费入口为 $49/month | 基于积分的 API 打包,高端代理会按倍数消耗 | 托管抓取、重试、渲染、住宅 / 移动池 | 高端请求会消耗更多积分 | 对想要结果、而不是采购代理的开发者,是强替代 |
各供应商公开定价并不均衡,因此本表强调入门信号和打包逻辑,而不是假装每个产品都能用一个简单的每 GB 数字直接比较。
[CP007, CP009, CP011, CP013, CP017, CP019]从图上看,Bright Data 在代理、工作流和合规层覆盖最宽;Oxylabs 在高端基础设施上紧咬,但在市场型产品或自下而上获客动作上显得没那么铺开。
[CP016, CP020, CP025, CP026, CP027, CP031]3.4 切换成本、多供应商并用、分发能力与供应获取
原始代理层的硬锁定效应低于很多企业软件品类,因为底层管道相对标准化。市场上的供应商普遍宣传 HTTP、HTTPS、SOCKS5、按用量付费或基于用量采购,以及试用或免费层,这降低了只需要传输和轮换的团队的切换摩擦。话虽如此,多供应商并用并非没有成本。一旦买方依赖 Web 解封锁工具、调好的地理定向、账户支持、特定工作流集成,或供应商托管的反机器人启发式规则,实际切换成本就会上升。分发能力也因类别而异。Bright Data 受益于围绕数据集、合规和企业级工具的很宽公共叙事;Apify 受益于 Actor 生态和 AI 工具分发。Oxylabs 的分发看起来最强的是企业采购和高端性能对话,而不是自下而上的免费层采用。不过,真实供给侧护城河的公开证据薄得多:供应商披露代理池规模和伦理来源主张,但不会暴露网络之间的重叠、独家来源合同或按买方队列的流失。这意味着分发和支持比任何可防御的供给独占性更容易被公开观察到。[CP017, CP020, CP022, CP028, CP029, CP030]
从公开证据看,Oxylabs 的护城河更像是在仍可多家并用的市场里执行高端路线,而不是牢不可破的锁定效应。
[CP028, CP029, CP031, CP037, CP038, CP039]3.5 护城河韧性、商品化压力与反向证据
反向证据并不是说 Oxylabs 弱,而是说这个品类结构上很难锁死。价格对比和排名来源显示,低端供应商可以用免费层、便宜数据中心池和低摩擦上手进攻。竞争对手撰写的对比和第三方对比也显示,托管 API 供应商可以把竞争话题完全从代理质量转向更简单、更高阶的数据采集结果。与此同时,Bright Data 这类高端同业还在持续扩展到数据集、浏览器自动化和智能体工具,这提高了原始代理和解封锁功能只成为更大工作流平台一个组件的风险。在这种环境里,Oxylabs 的护城河在企业买家重视高端性能、信任姿态和支持深度的地方最强;在买家优先考虑最低成本、最大自助简便性或集成工作流市场时更弱。因此,主要投资结论更细:Oxylabs 显然是品类领导者之一,但公开证据支持的是高端执行护城河,而不是垄断式锁定护城河;估值和流失假设应体现这一区别。[CP023, CP024, CP026, CP033, CP034, CP035]
| 护城河主张 | 威胁 | 严重性 | 公开证据 | 缓释措施 / 尽调问题 |
|---|---|---|---|---|
| 高端性能与支持 | 价值型供应商缩小质量差距,同时用更低价格切入 | 高 | Decodo 的最佳性价比定位;Webshare 和 IPRoyal 的低门槛产品 | 索取按买家细分和价格敏感度拆分的流失与赢单 / 输单数据 |
| 合规与信任基础 | Bright Data 也营销强认证和企业级控制 | 中 | Bright Data 合规页面和住宅代理页面;Oxylabs 信任叙事 | 验证 Oxylabs 是靠信任基础赢单,还是只是满足最低门槛 |
| 全栈产品广度 | Bright Data 进一步扩到数据集、浏览器、MCP 和智能体 | 高 | Bright Data vs Oxylabs 对比;Apify 对比页面 | 厘清 Oxylabs 路线图和各产品线附加率 |
| 代理网络规模 | 公开池规模说法无法说明独特性、重叠度或质量一致性 | 中 | AIMultiple 更看重 IP 声誉,而非原始规模 | 要求管理层提供来源集中度、重叠度和留存指标 |
| 企业 GTM | 托管 API 替代品可以完全绕过代理采购 | 高 | ScraperAPI 和 ScrapingBee 打包;Apify 工作流导向 | 衡量潜在客户有多常选择托管结果,而不是原始网络合同 |
| 标准协议带来的硬锁定较低 | HTTP/SOCKS 用法和试用机制让多供应商并行很容易 | 中 | Decodo、IPRoyal 和 Webshare 入门流程;ScraperAPI 抽象层 | 索取首次部署后的净收入留存和队列行为 |
本登记表把 Oxylabs 看似差异化的位置,与差异化可能主要只是竞争市场中高端执行的位置分开。
[CP022, CP028, CP029, CP031, CP037, CP038]3.6 要点
04财务情况
4.1 收入模型、定价架构与公开牵引力
Oxylabs 的公开财务图景以一个异常清晰的核心数字开始,随后很快变得不那么精确。公司 July 2026 投资公告称,Oxylabs 正以 $350 million ARR 运行,并获得全球 350,000 多个技术团队信任;官网仍称全球 15,000+ 客户信任它。这些数字都指向规模,但分母并不相同,因此应被视为强需求证据,而不是干净的客户队列模型。变现面清楚是经常性加用量的混合模式。公开产品页显示,住宅代理按 GB 销售,数据中心代理按 IP 和 GB 销售,Web Unblocker 按 GB 销售,Web Scraper API 的报价则可按流量或每 1,000 个结果销售,取决于套餐。这在经济性上很重要:Oxylabs 既不是纯席位制 SaaS,也不是一次性服务业务。它变现的是持续访问、性能和工作流便利性。这应该支撑经常性收入质量,但公开记录仍未披露产品线组合、折扣,或 ARR 中来自企业合同、自助渠道和收购渠道的比例。[CI001, CI004, CI005, CI006, CI007, CI008]
| 收入流 | 机制 | 单位 | 当前数值 / 状态 | 收入质量判断 | 尽调问题 |
|---|---|---|---|---|---|
| 住宅代理 | 轮换住宅访问及相关附加项带来经常性支出 | USD / GB | 公开套餐显示,入门约 $6/GB,高用量降至 $2.50/GB;提供免费试用 | 核心经常性用量收入,公开可见高端定价和量价折扣 | 按住宅代理队列拆分收入占比、流失和毛利率 |
| 数据中心代理 | 共享、独享和轮换数据中心产品按 IP 和按 GB 变现 | USD / IP 和 USD / GB | 公开定价按产品不同包括 $1.20/IP、$0.70/IP 和 $0.44/GB 入门价 | 降低对纯住宅支出的依赖,并可能拓宽买家组合 | 按 IP 与按流量数据中心收入的占比,以及企业支持附加率 |
| Web Unblocker | 按流量消耗销售的高端托管反爬访问层 | USD / GB | 公开套餐为 $9.40/GB 至 $7.50/GB,并包含速率限制和充值机制 | 若性能保持差异化,这可能是价值更高的收入流 | 毛利率、留存,以及 ARR 中来自 Web Unblocker 工作负载的比例 |
| Web Scraper API | 通过 API 套餐和结果 / 流量单位销售的托管结构化数据采集 | USD / GB 和 USD / 1K 结果 | 公开页面显示常规定价,以及按目标区分的结果定价和充值机制 | 把收入从原始连接上移到价值更高的工作流结果 | 抓取 API、自定义解析器和相关支持服务之间的收入拆分 |
| 搜索 / 快速 API 产品 | 面向实时 Web / AI 检索用例的搜索型 API 变现 | USD / 1K 请求 | Proxyway 显示 Oxylabs 参与快速和常规搜索 API 定价栈 | AI / 搜索需求可能在代理流量之外创造增量消耗上行 | 实际收入运行率、产品毛利,以及与核心抓取 API 收入的重叠 |
| 企业定制合同与交叉销售 | 定制打包、专属客户经理、支持、数据集、合规和收购渠道增售 | 合同 | 官方来源提到企业支持、收购、合作伙伴关系和 Fortune 500 / AI 客户,但未披露定价条款 | 可能是 ACV 最高的收入流,但公开透明度也最低 | ACV 区间、期限长度、折扣政策,以及收购客户群的扩张 |
| AI Studio | 基于积分的 AI 辅助数据提取工作流,按月度套餐打包 | 月度套餐和额度 | 公开套餐从 $12/month 起,向上扩展到定制企业包 | 在代理流量之外增加了更新、更高层的软件收入流 | 收入贡献、使用强度,以及相对核心代理产品的毛利率 |
本表区分公开可见定价的产品和透明度较低的企业打包。它看的是收入架构,不是已实现的 P&L。
[CI001, CI008, CI010, CI011, CI012, CI013]| 套餐 / 杠杆 | 公开标价 | 计费单位 | 包含容量 / 信号 | 仍未知事项 | 来源 |
|---|---|---|---|---|---|
| 住宅代理入门套餐 | ~$6/GB,5GB 月度套餐 | 按 GB | 入门点包含 24/7 支持和免费试用入口 | 实际折扣、年度条款,以及有多少用量按标价购买 | Oxylabs 住宅代理产品页 |
| 住宅代理企业套餐 | ~$2.50/GB,1TB 套餐 | 按 GB | 显示同一产品家族内存在明显规模折扣 | 实际混合住宅代理 ARPU 均值和客户用量组合 | Oxylabs 住宅代理产品页 |
| 数据中心代理 | 按变体不同,从 $1.20/IP、$0.70/IP、$0.44/GB 起 | 按 IP 或按 GB | 说明组合覆盖不同买家任务和成本结构 | 哪些数据中心 SKU 组合对收入和利润率最重要 | Oxylabs 数据中心产品页 |
| Web Unblocker | $9.40/GB、$8.60/GB、$7.50/GB 公开套餐 | 按 GB | 高端托管访问定价,配有速率限制和充值机制 | 用量集中度、企业折扣和毛利率画像 | Oxylabs Web Unblocker 页面 |
| Web Scraper API | 示例包括 $0.25/1K Amazon 结果、$0.50/1K Google 结果、$0.70/1K 其他结果,以及 $0.95/GB 充值 | 按 1K 结果和 GB | 确认定价逻辑同时绑定目标和使用画像 | 按目标类别的实际定价、解析器附加率和 SLA 条款 | Oxylabs Web Scraper API 页面 |
| Search API 对比项 | Proxyway 显示 Oxylabs 常规 search API 定价约 $1,另有提取加价项;其他供应商聚集在不同区间 | 按 1K 请求 | 有用的外部校验:Oxylabs 不只靠代理流量,也变现搜索型工作负载 | 搜索 / API 定价有多少是战略性亏损引流,有多少是利润中心 | Proxyway search APIs 2026 报告 |
| 市场底价校验 | Decodo 称住宅代理从 $2/GB 起,BestProxyCompare 还指出其他地方存在更低的混合价格底线 | 按 GB | 显示 Oxylabs 已发布高端价格带下方存在实质价格压缩风险 | Oxylabs 为守住更小或价格更敏感机会而打折的频率 | Decodo 定价和 BestProxyCompare 2026 排名 |
| AI Studio | $12 Starter;$62 Lite;$250 Standard;$1,200+ Custom | 按月 / 额度 | 显示 Oxylabs 直接变现更高层的 AI 工作流工具 | 积分消耗、超额使用模式,以及对核心代理买家的附加率 | Oxylabs AI Studio 定价 |
标价是公开的,但企业端实际经济性仍属私有。加入竞品行是为了显示变现压力,不是暗示产品完全相同。
[CI011, CI012, CI013, CI015, CI016, CI017]Oxylabs 把基础设施接入转成收入,路径从代理用量起步,再向托管 API、解锁能力和企业支持层上移。
[CI011, CI012, CI013, CI016, CI018, CI023]4.2 GTM 模式、单位经济代理指标与成本结构线索
公开证据显示,Oxylabs 把产品驱动漏斗机制和企业销售辅助结合起来。住宅代理、数据中心代理、Web Scraper API 和 Web Unblocker 都出现免费试用;付费计划也宣传专属客户经理、24/7 支持和面向更大用量的定制方案。这个组合意味着落地后扩张模式:买家可以快速测试,但随着支出上升,会被引导到接触度更高的商业关系。Webshare 收购强化了这一判断,因为交易后 Webshare 保留了不改价、自助、低成本的模式,让 Oxylabs 能服务开发者或较小买家,而不必彻底重塑高端 Oxylabs 品牌。成本侧只能看到碎片,但这些碎片很重要。Proxyway 称 Oxylabs 通过 Honeygain、直接 ISP 合作、应用开发者、JumpTask 和 SDK 获取住宅 IP;产品页强调高成功率、低延迟和全天候支持;影响报告强调 AI Studio、ScrapingBee 集成和信任基础设施投入。合在一起,这指向一个固定资产可能较轻,但在 IP 来源、工程、合规、支持和高端性能维护上 OPEX 强度不低的业务。[CI014, CI015, CI017, CI018, CI019, CI020]
| 指标 / 代理指标 | 公开数值 / 状态 | 置信度 | 重要性 | 证据基础 | 尽调问题 |
|---|---|---|---|---|---|
| 披露 ARR | ~$350M 年经常性收入(ARR) | 中 | 为任何估值或增长讨论提供收入顶线锚点 | Oxylabs 博客、TNW、Tesonet、EU-Startups、IBTimes | 董事会批准的按产品和季度拆分 ARR 桥表 |
| 官方客户基数定义 | 15,000+ 客户 vs 350,000+ 技术团队 / 客户 | 中 | 分母会实质改变对 ARPU、CAC 或留存的解读 | Oxylabs 首页 vs 2026 年 7 月融资报道 | 付费客户、技术团队、活跃账户和企业 logo 的精确定义 |
| 隐含估值 / ARR 倍数 | 按 $3.6B 估值和 $350M ARR 计算约 10.3x | 中 | 可作为高端基础设施估值的一阶基准 | Warburg / Oxylabs 融资事实,加管理层披露 ARR | 管理层对标准化 ARR 的看法,以及本轮融资使用的增长调整可比公司 |
| 粗略每外部实体 ARR | 按分母不同,每个技术团队约 ~$1K,每个客户约 ~$23.3K | 低 | 说明指标定义不清为何会成为重大尽调卡点 | 公开 ARR 加 350K vs 15K 的公开口径 | 按产品和账户类型拆分的队列级 ARPA/ARPU |
| 粗略每员工 / 专家 ARR | 按 400-500 名员工或 500+ 专家计算约 ~$700K–$875K | 低 | 为高端、类软件基础设施公司提供方向性的效率代理指标 | Proxyway 员工数估计和 Oxylabs「500+ experts」披露 | 当前员工数、全口径运营支出,以及按职能拆分的每 FTE 收入 |
| 毛利率驱动因素 | IP 来源、云 / 网络运营、反爬工程、支持、合规和客户管理 | 中 | 这些杠杆决定高端定价能否转化为耐久毛利 | Proxyway 来源描述,加官方支持 / 合规 / 产品页面 | 按产品拆分毛利率,并按来源、流量、支持和 R&D 拆分成本桶 |
| 销售效率代理指标 | 免费试用加专属客户经理,指向 PLG + 企业协助的混合模式 | 中 | 直接影响 CAC、回本周期和收入质量 | 官方定价 / 产品页面和 Webshare 收购后信息 | 按细分拆分 CAC、自助转化、销售周期长度和回本周期 |
| 公开留存指标 | 未披露 | 中 | 基础设施估值的耐久度高度依赖 NRR 和流失行为 | 保留的公开来源没有披露 NRR 或总流失 | 按产品拆分 NRR、logo 流失、扩张率和队列留存 |
公开指标缺失时,本表使用有边界的推断或定性代理指标,而不是假装已知精确经济性。
[CI001, CI006, CI007, CI008, CI009, CI018]可观察的单位经济逻辑从试用和产品使用开始,随后要么扩张到更高价值的托管产品,要么因价格敏感切换和未解决的流失而漏掉。
公开来源不披露 CAC、回本周期或 NRR,因此本图展示机制,而非具体利润率。
[CI014, CI015, CI019, CI021, CI023, CI025]4.3 资本充足性、融资依赖与资金用途
最清楚的融资事实也是战略上最重要的事实:Oxylabs 直到 July 2026 才引入外部资本。Warburg Pincus 通过 Capital Solutions Founders Fund 以约 $3.6 billion 估值投资 $130 million,官方和合作伙伴来源都把这轮融资描述为增长资本,而非救助融资。管理层明确把融资与扩大竞争优势、加速下一代产品、提升收购和合作能力相连。方向上这是利好,因为它暗示管理层把公司视为机会丰富,而非现金短缺。历史自举增长强化了这一解读。一家公司在首轮机构融资前已经达到数亿美元 ARR,与必须融资才能活下去的风险投资支持业务,不呈现同样的融资依赖画像。不过,公开资本充足性披露明显不完整。保留下来的公开来源没有披露手头现金、月度烧钱、现金跑道月数、债务额度、项目融资或 covenant 义务。因此,投资者可以得出融资依赖一直较低且已有新资本的结论,但仅凭公开证据还无法搭建严谨的现金模型或下一轮触发模型。[CI001, CI002, CI003, CI004, CI005, CI006]
| 项目 | 公开数值 / 状态 | 置信度 | 重要性 | 证据基础 | 尽调问题 |
|---|---|---|---|---|---|
| 手头现金 | 2026 年 7 月融资后未公开披露 | 中 | 现金余额决定真实战略灵活性和下行保护 | 留存的公开资料中,Oxylabs 或 Warburg 均未披露 | 交割后现金余额和最低运营现金政策 |
| 月度烧钱 | 未公开披露 | 中 | 烧钱速度决定这轮融资是在支撑进攻、防守,还是两者兼有 | 没有公开烧钱速度拆解 | 按运营费用类别拆分的月度烧钱速度,并对比融资前后 |
| 现金跑道月数 | 未公开披露;历史上自举经营叠加 $130M 新资金,方向上改善了现金跑道 | 中 | 现金跑道决定融资依赖度和并购能力 | 仅有自举经营历史和融资规模 | 基准、悲观和并购情景下的内部现金跑道模型 |
| 计划资金用途 | 延长竞争优势、加速下一代产品,并扩大收购 / 合作 | 中 | 资金用途能看出资本是用来扩张,还是弥补弱点 | Warburg、Oxylabs、EU-Startups、TNW、Tesonet 等来源 | 研发、GTM、收购和资产负债表之间的详细资本分配计划 |
| 下一轮触发条件 | 未公开说明;除非管理层推进规模明显超过内部现金流承受能力的并购或 AI 基础设施扩张,否则可能不是刚需 | 低 | 有助于判断当前估值能否在没有短期融资压力的情况下守住 | 来自首轮外部融资和自举经营历史的推断 | 董事会对可接受杠杆、流动性和下一轮融资阈值的看法 |
| 债务 / 项目融资义务 | 留存来源未识别出公开债务额度、库存融资或项目融资义务 | 中 | 看不到结构性债务,有利于更干净的现金流叙事,但未经确认不能直接假设 | 已审阅来源中没有债务相关表述 | 债务明细、留置权、担保和表外承诺 |
| 融资依赖历史 | 历史上低:运营十年后才在 2026 年 7 月首次接受外部投资 | 高 | 说明公司在没有反复外部融资的情况下活下来并扩大了规模 | Warburg、Oxylabs、TNW、EU-Startups、Tesonet 等来源 | 历史现金流量表和盈利时间线 |
历史轮次时间线放在公司概况;本表只关注资本充足性、依赖度,以及 2026 年融资如何改变现金叙事。
[CI002, CI003, CI004, CI005, CI006, CI033]公开记录给出了清晰的 ARR 和估值核心数字;但客户和员工分母定义不一致,只能粗略推导单位经济区间。
[CI001, CI006, CI008, CI009, CI024, CI037]Oxylabs 结构上更像 OPEX 密集,而非 CAPEX 密集:客户现金似乎用于来源采购、流量交付、支持、合规和产品扩张;2026 年股权融资主要放大战略选择权。
没有公开现金流量表;本图只反映可见成本和资金用途类别。
[CI003, CI005, CI021, CI023, CI025, CI033]4.4 财务判断、估值输入解读与尽调阻塞项
当前公开记录支持一个正面但尚未被完整数据撑住的财务判断。正面很强:管理层报告的 $350 million ARR、已验证的 $130 million 增长轮、$3.6 billion 估值、广泛的经常性变现面、可见的高端定价,以及公司在接受机构资金前能自举超过十年的证据。这些都是收入韧性和商业模式可行性的高质量信号。负面在于不透明。Oxylabs 不披露按产品或客户细分的收入组合、各流毛利率、实际折扣、CAC、回本周期、NRR、烧钱、现金跑道或债务。15,000 客户与 350,000 技术团队之间的不一致,也让没有管理层定义的粗略单位经济测算用处有限。TNW 又加了一层反向因素:公司所在的网页数据品类围绕抓取和住宅来源仍带有法律与声誉包袱,可能转化为采购摩擦和持续合规成本。因此,财务结论是,Oxylabs 看起来像一家高端、资本较轻的基础设施软件公司,收入质量扎实,但投资者仍需要私有经营数据,才能有信心判断利润率、现金效率和当前估值倍数的耐久性。[CI009, CI028, CI029, CI031, CI032, CI033]
| 缺失的私有指标 | 对承销判断的影响 | 精确尽调路径 |
|---|---|---|
| 按产品和客户分部拆分的收入结构 | 没有收入结构,投资者无法判断 ARR 有多少来自高价解封 / API 业务,有多少只是更商品化的代理流量 | 要求提供住宅代理、数据中心代理、解封器、爬虫 API、搜索 API、数据集和并购渠道之间的 ARR 与总收入拆分 |
| 实际定价和折扣 | 公开标价可见,但企业客户的实际经济性和竞争性折扣不可见 | 要求提供前 50 大合同、价格手册、折扣授权规则,以及按产品拆分的净收入实现率 |
| 按业务流拆分的毛利率 | 只有在资源获取、流量和支持成本之后仍留下可持续毛利,高价定价才有价值 | 要求按产品族提供毛利率桥,并拆出资源获取、云、支持和合规成本 |
| 获客成本(CAC)、回本周期和销售周期数据 | 没有获客成本和转化数据,无法承销 PLG 加企业协助销售模式 | 要求提供自助转化漏斗、现场销售 CAC、回本周期,以及按分部拆分的销售周期长度 |
| NRR、logo 流失和扩张行为 | 经常性基础设施估值高度依赖队列耐久度和扩张机制 | 要求提供总留存、NRR、按产品扩张和流失原因的队列表 |
| 现金、烧钱速度、现金跑道和债务明细 | 仅靠公开记录无法可信建模资本充足性 | 要求提供月度现金余额、烧钱预测、债务与契约明细,以及下行情景现金跑道分析 |
| 15K 客户与 350K 技术团队的指标定义 | 除非精确定义这些口径,否则粗略 ARPU 和单位经济估算并不可靠 | 要求提供客户、委托客户、团队、活跃账户、付费 logo 和企业 logo 的指标字典 |
这些缺口不是表面问题;它们直接决定 Oxylabs 的高价 ARR 和估值主张能否转化为站得住的投资模型。
[CI009, CI031, CI032, CI033, CI034, CI038]4.5 要点
05产品与技术
5.1 产品定义、模块范围与客户工作流
Oxylabs 的产品面已经不只是“买代理”。官网和产品页展示的栈从住宅和数据中心连接开始,然后上升到 Web Unblocker、Web Scraper API、Fast Search API,以及 AI-Scraper 等 AI Studio 应用。放到工作流里,这意味着客户可以在直接流量控制、托管请求执行、特定领域结构化抽取,或面向 AI 管线的自然语言驱动数据采集之间选择。官方文档把预期递进说得很明确:Web Scraper API 快速上手文档从一次 API 请求开始,隐藏代理轮换、渲染和反机器人逻辑;Web Unblocker 用代理端点加控制 headers 包住高难目标;AI-Scraper 再往上一层,让用户用自然语言描述想要的输出。因此,最重要的产品结论是架构性的:Oxylabs 卖的是越来越多的抽象。这有利于客户采用,因为它对应不同待完成任务——从基础设施很重的抓取团队,到更关心输出而非低层网络决策的开发者和 AI 构建者。[CE001, CE004, CE005, CE006, CE008, CE010]
| 模块 / 资产 / 产品线 | 主要用户 | 状态 / 成熟度 | 差异化 | 尽调缺口 |
|---|---|---|---|---|
| 住宅代理 | 爬虫团队、SEO、电商、AI 数据运营方 | 成熟 | 大规模 IP 池、地理位置控制、高级支持、自适应过滤 | 按目标类型拆分的实际表现,以及按地区拆分的毛利率仍属私有 |
| 数据中心代理 | 高流量技术用户和成本敏感的结构化爬虫团队 | 成熟 | 响应时间快、多种计费模式、支持多协议 | 需要更清晰披露共享、专用和轮换需求之间的结构 |
| Web Unblocker | 抓取受保护网站、但不想自建绕过逻辑的团队 | 成熟 | 托管指纹、重试、地理 / 渲染 / 会话控制 | 覆盖更多目标的独立同口径成功率基准仍然有限 |
| Web Scraper API | 想要结构化数据、但不想管理基础设施的开发者和数据团队 | 成熟 | 结合代理轮换、渲染、解析和受支持目标 | 需要专用解析器和自定义用例的附加率数据 |
| Fast Search API | AI 应用、智能体和高吞吐 SERP 用户 | 增长期 | 面向应用或 LLM 工作流的低延迟搜索结果 | 公开定价和客户采用细节少于核心产品 |
| AI Studio / AI-Scraper | 想用提示词抽取数据的开发者和 AI 构建者 | 新兴 | 自然语言抽取、模式生成、JSON/Markdown、SDK | 需要证明企业采用、准确率和经济性优于经典 API |
| 文档 / SDK / GitHub 层 | 评估或集成平台的开发者 | 增长期 | 指南、文档、SDK 和快速上手流程降低集成摩擦 | 需要使用数据证明开发者工具对转化和留存的影响有多大 |
成熟度标签是基于可见文档深度、定价清晰度、支持目标和外部基准作出的证据型判断,而不是内部路线图披露。
[CE004, CE008, CE010, CE011, CE013, CE015]| 用户任务 | 当前工作流 | Oxylabs 方案 | 可衡量收益 | 局限 |
|---|---|---|---|---|
| 收集本地化公共网络原始数据 | 手动配置代理并管理轮换 | 住宅或数据中心代理 | 用高质量 IP 和地理覆盖提供更底层的控制 | 用户仍需自己负责解析和工作流逻辑 |
| 抓取有反爬系统保护的高难度目标 | 手动组合代理、无头浏览器和重试 | Web Unblocker | 把解封复杂度藏在一个代理端点和请求头后面 | 高价定价和部分黑箱行为 |
| 收集特定目标的结构化数据 | 内部搭建爬虫、解析器和调度器 | Web Scraper API | 通过一个 API 返回结构化或可解析数据 | 覆盖范围和解析器深度因目标而异 |
| 为 AI 跑高吞吐搜索检索 | 调用搜索引擎,或用自建工具抓取 SERP | Fast Search API | 为应用和 LLM 大规模提供低延迟 SERP 检索 | 公开功能和定价细节少于旗舰产品 |
| 用自然语言从任意页面抽取信息 | 手写选择器、模式和解析器维护逻辑 | AI-Scraper | 用提示词抽取,并输出 JSON 或 Markdown | 模型驱动抽取可能更新,也比经典解析器更难做基准 |
| 快速原型或集成 | 阅读通用 API 文档并从零搭建 | 快速上手、文档、SDK、GitHub 指南 | 开发者上手更快,集成摩擦更低 | 文档质量本身不能保证生产可靠性 |
本表按客户任务而不是 SKU 标签组织,因为 Oxylabs 越来越靠抽象层级竞争,而不只是靠 IP 类型。
[CE001, CE004, CE005, CE008, CE010, CE011]客户可从原始代理一路进入更高抽象层,直到提示词驱动的提取。
[CE001, CE004, CE008, CE010, CE011, CE016]5.2 架构、部署模型、集成与关键依赖
公开文档显示,Oxylabs 产品有共同运行模式:客户输入通过 API 或代理端点进入,Oxylabs 选择或轮换网络路径,反机器人逻辑和渲染在后台处理,数据以原始内容、结构化 JSON 或 AI 形态输出返回。Web Scraper API 暴露 sources、URL 或 query 输入、parse flags、render modes、geo-location,以及 realtime 和 asynchronous delivery 等不同集成方式。Web Unblocker 暴露用于地理选择、渲染、截图、sticky sessions 和强制自定义 headers 的 headers,同时在幕后使用重试和浏览器指纹逻辑。AI Studio 用 prompts、schemas、SDKs 和面向 MCP 的集成做了更多抽象。这个架构很强,但也让 Oxylabs 依赖多个外部系统:住宅 IP 供给、云和网络运营、目标网站的反机器人反制、受支持的目标专用解析器,以及开发者能够信任的文档或 SDK 层。Proxyway 关于代理 API 的研究强化了这一点:这个品类本质上是多层协调问题,而不是简单代理列表。[CE005, CE006, CE007, CE008, CE009, CE012]
| 层级 / 流程 / 组件 | 角色 | 依赖项 | 风险 |
|---|---|---|---|
| 持有凭证的 API 用户和代理端点 | 认证访问,并区分产品权限 | 控制台、凭证发放、用户管理 | 配置错误或薄弱上手流程会拖慢采用 |
| 输入抽象层 | 接收 URL、查询、目标 ID、提示词、模式、请求头和地理指令 | 文档清晰度、SDK、特定目标语法 | 参数面过于复杂会抬高用户错误率 |
| 网络和路由层 | 选择 IP 类型、轮换路径、管理重试,并让请求本地化 | 住宅供应、数据中心机群、地理覆盖、流量管理 | 来源质量恶化或地理区域被封会拉低表现 |
| 反爬 / 渲染层 | 处理 CAPTCHA、浏览器指纹、JavaScript 渲染和目标适配 | 无头 / 浏览器技术、ML 规则、目标监控 | 防守方持续进化;回归会伤害成功率 |
| 解析和结构化输出层 | 把原始页面转成结构化 JSON 或 AI 形态输出 | 特定目标解析器、模式逻辑、提示词保真度 | 解析器漂移或幻觉式抽取会削弱信任 |
| 交付和集成层 | 通过 API 和 SDK 返回原始 HTML、JSON、截图或 AI 输出 | 实时 / 推拉基础设施、云交付、SDK 维护 | 集成中断或文档缺口会拖慢采用 |
| 支持和信任层 | 增加企业支持、合规姿态和问题解决能力 | 24/7 团队、信任流程、安全控制 | 如果支持或信任证据滑坡,高端定位会变弱 |
Oxylabs 的架构最好理解成托管请求编排系统,而不是静态代理列表。
[CE005, CE006, CE007, CE008, CE012, CE017]Oxylabs 的产品栈把原始接入、托管编排、解析、AI 提取和信任控制叠成一个公开 Web 数据平台。
[CE005, CE006, CE008, CE011, CE017, CE024]产品依赖外部供给、目标网站行为和面向开发者的集成界面,Oxylabs 必须持续协调这些变量。
[CE017, CE022, CE025, CE029, CE036]5.3 差异化、产品成熟度与路线图信号
Oxylabs 显然有广度,但公开来源也显示,单靠广度已经不再独特。Bright Data 营销同样完整的解封锁、渲染、CAPTCHA 处理和浏览器级控制。ScraperAPI 提供高端住宅和移动池、渲染、受支持参数,甚至为困难目标提供超高端路由。Apify 从另一个角度竞争,它提供市场式抓取工作流;价值型供应商也在持续缩小原始代理访问差距。因此,差异化问题更微妙。Oxylabs 最强的公开理由是,产品家族分层清晰:原始代理、代理 API、结构化抓取器、快速搜索、AI 工具和企业支持,都指向同一个任务——规模化、可靠地采集公开 Web 数据。产品成熟度信号也以有用的方式不均衡。住宅和数据中心代理看起来成熟;Web Scraper API 和 Web Unblocker 也成熟,有快速上手文档、产品文档、大量受支持目标、明确参数和外部基准测试。AI Studio 感觉更新:它已经有定价、SDKs 和文档,但显然被定位为仍在扩展企业相关性的新抽象层。[CE010, CE011, CE013, CE018, CE019, CE020]
| 日期 / 阶段 | 功能 / 里程碑 | 状态 | 含义 | 来源 |
|---|---|---|---|---|
| 2025 | AI Studio 发布 | 已发布 | 在核心 API 之上引入 AI 原生抽取和搜索层 | Impact 报告 |
| 2025 | ScrapingBee 收购 | 已完成 | 扩大开发者和爬虫 API 覆盖 | Impact 报告和融资博客 |
| 2026 | Fast Search API 的 AI / 智能体定位 | 活跃 | 说明产品在适配 AI 搜索和 LLM 工作流 | Fast Search API 页面和 Proxyway 搜索研究 |
| 2026 | AI Studio 定价和文档上线 | 活跃 | 表明商业化已超过实验阶段 | AI Studio 定价和文档 |
| 2026 | 开发者快速上手在 1-2 个月内更新 | 活跃 | 显示文档维护和上手投入仍在持续 | 开发者文档页面 |
| 2026 | Web Unblocker 和 Web Scraper API 仍接受外部基准测试 | 活跃 | 核心托管产品在竞争定位中仍然重要 | Proxyway 和 Bright Data 对比内容 |
公开路线图来自产品发布、文档新鲜度和商业化界面的推断,而不是正式路线图披露。
[CE010, CE011, CE013, CE016, CE018, CE020]核心连接和托管抓取产品看起来已经成熟;AI Studio 较新,但战略重要性高。
[CE010, CE011, CE013, CE018, CE019, CE020]5.4 信任、质量、安全与合规控制
产品质量和信任控制是产品的一部分,不是旁支。官方文档和产品页反复强调高成功率、快速响应时间、地理定向、重试、解析器可靠性和 uptime 等运营质量信号;文档层也加入了身份认证、速率限制、错误码、渲染开关和会话保持等具体控制。Web Scraper API 文档明确提到包括 SOC 2 Type II 合规在内的企业安全标准;更早的公司来源也通过 Trust Center、合规人员配置和伦理来源姿态强化了更宽的信任叙事。这些控制很重要,因为 Oxylabs 所在品类会被目标网站、监管机构和采购团队一起审视数据访问方式。不过,一些信任主张是准入门槛,而不是差异化。Bright Data 同样宣传 CAPTCHA 处理、指纹识别和重合规话术,第三方对比也持续把多个托管栈排为可用于困难网站。因此,真正的产品质量问题不是 Oxylabs 是否有控制,而是随着时间推移,这些控制能否比替代方案更可靠、更适合企业使用,也更容易获得支持。[CE002, CE003, CE008, CE009, CE018, CE019]
| 控制 / 认证 / 质量指标 | 状态 | 范围 | 缺口 |
|---|---|---|---|
| SOC 2 Type II 表述 | 公开声称 | Web Scraper API 文档提到企业安全标准 | 需要每个产品的范围边界和审计时间新鲜度 |
| 通过专用 API 用户认证 | 可见 | 文档将产品 API 凭证与控制台登录分开 | 需要更深入公开说明角色粒度和企业管理 |
| 速率限制和错误码 | 可见 | 文档展示 401、429、550 以及并发 / 速率限制行为 | 需要按产品和客户层级公开 SLO |
| 地理定位和会话控制 | 可见 | Web Unblocker 和爬虫文档暴露位置和会话参数 | 需要明确各产品的位置覆盖质量 |
| 渲染和截图控制 | 可见 | Web Unblocker 文档暴露 html/png 模式和强制请求头 | 需要公开与竞品相比的渲染可靠性基准 |
| 信任 / 伦理姿态 | 在公司多个公开界面可见 | 前文报告章节确认了 Trust Center、合规和伦理表述 | 需要更清楚公开映射信任主张如何落到产品级执行 |
| 性能质量信号 | 可见,但部分为自报 | 成功率、正常运行时间、响应时间主张出现在产品页和对比页 | 需要长期、标准化的独立性能报告 |
可见控制相当充分,但很多仍更适合解读为信任信号,而不是公开记录中完全可审计的证明。
[CE003, CE009, CE018, CE019, CE024, CE025]5.5 要点
06客户情况
6.1 客户分层、买方组合与采用面
公开客户记录显示,Oxylabs 服务的是分层客户群,而不是单一狭窄细分。公司官网称全球 15,000+ 客户信任它;July 2026 融资材料则把口径扩大到全球 350,000 多个技术团队。案例研究中心还称,客户基础横跨多个行业的中端市场到企业组织;案例材料显示,它在零售情报、SEO 分析和公共部门执法中有真实使用。这个组合意味着几类不同的买方 / 用户 / 付款方组合:企业数据或运营团队采购高端代理和抓取基础设施,软件公司把 Oxylabs 嵌入自己的客户产品,公共机构或使命驱动项目用定制爬虫支持执法或研究。渠道多样性也由此可见。Oxylabs 以高端品牌直接销售,但收购 Webshare 后保留了一条摩擦更低的自助路径,用来服务预算敏感或开发者主导的买家。主要限制是定义清晰度:公开来源没有解释 “clients”、“tech teams” 和 “customers” 的精确差异,因此能观察到细分方向,但各细分收入权重仍不透明。[CU001, CU002, CU003, CU011, CU012, CU023]
| 分群 | 买方 / 用户 / 付款方 | 用例 | 规模信号 | 收入 / 战略价值 | 缺口 |
|---|---|---|---|---|---|
| 企业零售情报和商业软件 | 企业软件供应商内部的运营、数据和平台团队 | 高频价格、目录和零售数据采集 | Wiser 案例研究;15,000+ 客户和企业评论措辞 | 可能是高价值经常性账户,并且对正常运行时间要求高 | 公开来源未显示 ACV 或合同期限 |
| SEO / 搜索情报供应商 | SEO 软件运营方及其终端用户 | SERP 和搜索引擎数据管道 | Epicup 案例研究和爬虫 API 产品动作 | 验证其用于经常性、面向客户的数据产品 | 公开来源未显示类似 SEO 账户数量 |
| 公共部门 / 使命驱动机构 | 部门负责人和官员 | 监控环境违规和线上执法 | EPD 案例研究,且每周持续使用爬虫 | 战略上重要,可证明工作流影响和声誉合法性 | 不清楚关系是付费、补贴还是完全公益 |
| AI / 智能体构建者和开发者 | ML 工程师、产品团队、开发者 | 事实锚定、搜索、提示词抽取、AI 数据管道 | 350,000+ 技术团队表述和 AI Studio 界面 | 可能是较新产品的快速增长队列 | 公开没有拆分 AI 买方和传统爬虫买方的采用情况 |
| SMB / 自助和预算敏感用户 | 创始人、个人运营者、小型技术团队 | 基础代理需求、测试或较低复杂度爬取 | Webshare 渠道和混合评论平台证据 | 提供更宽的漏斗和低摩擦入口 | 可能更价格敏感,也更容易流失到更便宜替代品 |
| 企业导向评论平台用户 | IT、竞争情报和数据基础设施评估者 | 高流量网络情报采集 | G2、Gartner、Proxyway 和 Capterra 的企业语境 | 支撑采购流程中的高端品牌可信度 | 评论界面没有暴露分部收入贡献 |
分群按可观察的客户任务和证明类型归类,而不是按内部 CRM 类别;后者仍属私有。
[CU001, CU002, CU003, CU011, CU017, CU023]| 指标 | 数值 | 日期 | 来源 | 置信度 | 含义 | 缺失分母 |
|---|---|---|---|---|---|---|
| 全球客户 | 15,000+ | 2026-07 | Oxylabs 首页 | 中 | 确认装机基础广、品牌触达大 | 未披露付费 logo 定义 |
| 全球技术团队 / 客户 | 350,000+ | 2026-07 | Oxylabs 融资材料 | 中 | 提示漏斗顶部很大,或长尾使用覆盖面很广 | 不清楚是否包含免费、间接或非付费用户 |
| 案例研究覆盖面 | 从中端市场到企业客户,横跨多个行业 | 2026-07 | Oxylabs 案例研究中心 | 中 | 显示产品能够跨垂直领域适配 | 未按行业或地区披露数量 |
| Wiser 运营节奏 | 每天在数千个域名上多次采集数据 | 当前案例研究 | Wiser 案例研究 | 中 | 企业零售工作流中的生产使用信号很强 | 未披露支出或席位数 |
| EPD 爬虫吞吐量 | 每周扫描 ~400 条广告 | 自 Jul. 2024 起 / 展示 2025 结果 | EPD 案例研究 | 中 | 证明运营层面的重复使用和影响 | 不是典型商业部署分母 |
| EPD 执法结果 | 发起 15 起调查;确认 4 起违规 | 集成之后 | EPD 案例研究 | 中 | 显示部署带来可衡量的工作流影响 | 未直接关联收入或留存 |
| 评论平台规模 | 421 条 G2 评论;22 条 Capterra 评论;20 条 Gartner 评论 | 2026-07 快照 | 评论平台 | 中 | 显示有意义的评论足迹和广泛使用历史 | 评论数量不等于客户数量 |
该表混合了客户数量、案例研究用量和评论平台规模,因为公开来源没有给出一张干净的采用情况总览。
[CU001, CU002, CU003, CU004, CU009, CU010]从发现到嵌入式工作流使用,梳理客群、采用触点和扩张循环。
这些阶段对应公开可见的获客到扩张路径;公开资料未披露各阶段之间的转化率。
[CU003, CU014, CU015, CU022, CU027]从广泛认知到嵌入式重复工作流使用的示意路径。
百分比是示意性代理指标,基于广泛评论 / 营销触达与少量直接具名生产案例之间的落差。
[CU001, CU002, CU011, CU021, CU030]6.2 具名客户证据、部署成熟度与结果具体性
具名客户证据足以支撑商业叙事,即便还不足以支撑完整留存模型。Wiser 看起来是生产级企业参考,而不是试点。其案例研究描述了一个数据密集的零售工作流,每天多次访问数千个域名,并把 Oxylabs 数据中心代理称为 Wiser 运营的关键部分。Epicup 是另一个可信的生产型参考:这家 SEO 软件供应商称,Oxylabs 抓取工具帮助它在满足定价、延迟和成功率要求的同时,保持高容量、新鲜 SEO 数据。Environmental Protection Department 案例在战略上不同。它是公共利益或公益合作,而不是经典经常性收入证明,但展示了部署深度和可量化运营结果。该爬虫自 July 2024 起已集成进日常运营,每周扫描约 400 个广告,并支持了多项调查和已确认违规。合在一起,这些例子说明 Oxylabs 不只是被好奇用户试用;它已经嵌入带有经常性运营后果的真实工作流。限制在于广度:相对公司声称的整体客户基础,公开具名证据集合仍然很小。[CU004, CU005, CU006, CU007, CU008, CU009]
| 客户 | 细分市场 | 部署 / 使用场景 | 生产还是试点 | 结果 | 局限 |
|---|---|---|---|---|---|
| Wiser Solutions | 零售情报 / 商务执行 | 数据中心代理,支撑跨数千个域名的高频零售数据采集 | 类生产 | 高质量零售数据持续流入;运营具成本效率;Wiser 扩张时,双方预期继续合作 | 未量化支出、留存或合同期限 |
| Epicup | SEO 软件 | SERP Scraper API / Web Scraper API,为终端用户提供实时 SEO 数据 | 类生产 | 成功率高、响应时间短、定价公平,SEO 洞察更新鲜也更可靠 | 单一供应商引述;没有独立 ROI 审计 |
| Lithuania Environmental Protection Department | 公共部门 / 执法 | 面向线上环境违规的定制广告站点网页爬虫 | 生产 / 已部署,但不是典型收入证明 | 每周扫描 ~400 条广告,发起 15 起调查,确认 4 起违规,人工监测减少 | 可能是公益或使命驱动合作,而非普通订阅经济 |
三个引用都显示了真实工作流部署;主要差异在于它们证明的是商业复购,还是使命驱动影响。
[CU004, CU005, CU007, CU009, CU026, CU030]具名案例研究、评论平台与宽泛总客户数声明之间,证据质量差异很大。
这些序数质量评分概括各类证据对投资判断的帮助程度。
[CU004, CU007, CU009, CU013, CU030]6.3 留存可见度、重复使用与满意度证据
公开满意度证据有意义但不完整。G2 有大量评论,评分 4.5,用户评论反复强调可靠性、支持响应和易设置。Capterra 在服务、设置和支出透明度上给出类似正面图景,同时也暴露了高端定价抱怨,以及部分一人运营者被拒绝免费试用的事实。Trustpilot 明显更复杂:近期多条评论称赞支持,但退款摩擦、账户问题和支持体验不一致也出现。Gartner 提供企业视角:评论者认为 Oxylabs 对严肃、高容量、企业规模抓取有价值,同时指出较小运营者可能在别处找到更高性价比。Thunderbit 的 2026 评测综合了同样的分裂,认为企业买家和较小运营者对平台的体验不同,定价和 KYC 是常见摩擦来源。综合判断很清楚:Oxylabs 可能在更大、更技术成熟的客户中拥有强重复使用,但公开来源仍没有披露 GRR、NRR、续约率、合同长度或队列行为。满意度看得见;耐久性看不见。[CU013, CU014, CU015, CU016, CU017, CU018]
| 指标 | 数值 / 空值 | 细分市场 | 置信度 | 尽调请求 |
|---|---|---|---|---|
| G2 评分 | 421 条评论,4.5 / 5 | 用户基础广,包括企业评估者 | 中 | 按产品和客户规模拆分评分 |
| Capterra 评论调性 | 安装、支持、稳定性和支出透明度评价正面;也能看到价格投诉 | SMB 到企业客户混合 | 中 | 按套餐和公司规模提供原始评论分布 |
| Trustpilot 评分 | 3.9 / 5,支持和退款体验评价不一 | 混合 / 长尾公开用户 | 中 | 将投诉类别映射到细分市场和生命周期阶段 |
| Gartner 企业适配度 | 企业用户看到了价值;小客户可能在别处找到性价比更高的选择 | 企业评估者 | 中 | 按细分市场提供续约和 NRR,以验证该分化 |
| 公开 NRR / GRR | 未披露 | 所有细分市场 | 中 | 提供 NRR、GRR、logo 流失和队列留存 |
| 公开合同期限 / 续约率 | 未披露 | 企业和中端市场 | 中 | 提供期限组合、续约率和扩张行为 |
| 案例研究中的重复使用代理指标 | Wiser 和 EPD 工作流显示运营层面的重复使用 | 仅具名验证 | 低 | 提供具名客户的产品使用频率和留存 |
评论平台能揭示满意度,但看不到持久经济性;具名客户验证能揭示重复使用,但看不到商业留存。
[CU013, CU014, CU015, CU016, CU017, CU018]按客群示意留存衰减,公司未披露实际数据。
留存值是示意占位(0-100),用公开满意度和价格敏感信号框定不同客群的耐久性差异;Oxylabs 未披露实际队列留存数据。
[CU017, CU018, CU021, CU028]6.4 扩张循环、集中度风险与采购摩擦
公开证据支持一个可信的落地后扩张模式,但没有量化。产品栈本身暗示了强扩张路径:客户可以从原始代理开始,转向 Web Scraper API 或 Web Unblocker,之后再采用 Fast Search API 或 AI 原生工具。具名案例研究也体现了工作流深度,而非一次性购买,尤其是 Wiser 的多域名经常性采集和 Epicup 嵌入式 SEO 数据管线。与此同时,风险也不小。Thunderbit、Trustpilot、Gartner 和 Capterra 都强化了同一点:高端定价和上手或 KYC 摩擦,会成为较小用户或不需要企业级服务买家的障碍。BestProxyCompare 和竞争者定价界面显示,低成本替代方案仍很多,这意味着 Oxylabs 客户基础的一部分耐久性可能取决于持续的性能领先和支持质量,而不是低价。公开证据在集中度上最弱。保留下来的来源没有披露头部客户敞口、合同期限、续约队列、合作伙伴依赖或按细分划分的收入组合。因此,商业判断是平衡的:Oxylabs 看起来有可信扩张向量和广泛采用,但集中度和留存经济性仍是重大尽调缺口。[CU022, CU023, CU024, CU025, CU027, CU028]
| 扩张驱动因素 | 集中风险 | 影响 | 尽调路径 |
|---|---|---|---|
| 更宽的产品栈 | 客户可从代理起步,再扩展到 API、解锁器和 AI 工具 | 如果增购真实存在,NRR 有上行潜力 | 按队列索取附加率和产品采用阶梯 |
| 跨垂直领域使用场景 | 零售、SEO、AI 和公共部门验证显示适用面较广 | 表面上降低单一垂直领域依赖 | 索取按垂直领域和地域拆分的收入结构 |
| 高端企业定位 | 可能加深与大客户的关系 | 若企业客户占比过高,可能提高头部客户集中度 | 索取前 10 和前 20 客户收入集中度 |
| Webshare 自助渠道 | 扩大漏斗,也降低进入摩擦 | 也可能稀释支持资源,或把低价值用户推入长尾 | 按渠道索取获客、激活和转化指标 |
| KYC / 入门摩擦 | 可能拖慢或劝退小买家 | 提高中非企业细分市场的流失或丢单风险 | 索取丢单原因和入门流程流失指标 |
| 市场中的低价替代品 | Webshare、Decodo 和其他低成本选项仍然显眼 | 压制定价权和小客户留存 | 按账户规模和竞争对手索取流失分析 |
公开来源支持扩张逻辑和流失压力,但没有给出量级。
[CU022, CU023, CU025, CU027, CU032, CU033]6.5 要点
07风险
7.1 监管、隐私和使用合规风险
Oxylabs 的法律姿态比外界对代理行业的刻板印象更稳,但它能活动的边界正在收紧。公司把合规做得异常外显:信任中心、适用于所有客户的 KYC 政策、通用条件协议、可接受使用政策和数据处理协议。这些确实能降风险,尤其是可接受使用政策明说,自动化数据采集并非天然违法,但会因方法和使用场景不同触碰法律红线。问题在于,外部监管背景仍在变严。GDPR 已经约束个人数据处理;EDPB 2026 年 7 月的网页抓取指引也说得很清楚:被抓取数据含有个人数据时,信息公开可得并不会免除隐私义务。与此同时,EU AI Act 下 Article 50 透明度规则将于 2026 年 8 月 2 日起适用于大量 AI 交互和 AI 生成输出;Oxylabs 正在推进面向 AI 的数据抽取界面,这一点很关键。判例法仍然细腻,并非全面放行:Meta v. Bright Data 在一个语境下支持公开、未登录抓取,但 LinkedIn 条款仍明确禁止抓取和自动化采集。因此,真正的法律风险不是今天已经看见某个针对 Oxylabs 的执法行动,而是目标站点特定条款、隐私法和 AI 义务共同制造出不断移动的合规负担,客户或监管者未来可能更激进地测试这条边界。[CR001, CR002, CR003, CR004, CR005, CR006]
| 规则 / 案件 | 司法辖区 | 状态 | 可能性 | 严重性 | 缓释措施 | 残余风险 | 尽调路径 |
|---|---|---|---|---|---|---|---|
| GDPR + EDPB 针对个人数据网页抓取的指南 | 欧盟 | 现行;EDPB 03/2026 于 July 2026 通过并征求意见 | 中高 | 高 | KYC、信任控制、客户审查、DPA、隐私和伦理姿态 | 高,因为许多公开网络使用场景会触及个人数据,并需要最小化处理或合法利益分析 | 审查合法依据手册、客户使用场景筛查、数据最小化控制,以及各产品的控制者 / 处理者分工 |
| 目标站点条款和反抓取合同风险 | 全球 / 具体端点 | 当前存在,且高度依赖具体事实 | 高 | 高 | AUP 承认不同方法和使用场景有各自限制;KYC 和滥用监测降低鲁莽用法 | 高,因为即便部分公开数据判例有利,合法性仍会因端点和客户工作流而异 | 测试头部目标域名的法律审查、产品防护,以及拦截高风险客户使用场景的流程 |
| EU AI Act 第 50 条关于 AI 输出和交互的透明度义务 | 欧盟 | 对适用系统自 2 Aug 2026 起适用 | 中 | 中高 | 可通过产品设计和文档落地 | 中,因为 Oxylabs 正在扩展到面向 AI 的提取界面,可能附带额外标注或披露义务 | 梳理哪些面向 AI 的产品触发第 50 条,并验证机器可读标记 / 披露控制 |
| 客户滥用、制裁或不道德使用场景 | 多司法辖区 | 持续 | 中 | 中高 | 强制 KYC 问卷、拒绝服务权、AUP 暂停 / 终止权 | 中,因为代理基础设施中的滥用风险永远不会消失 | 索取客户拒绝率、制裁筛查、升级路径和定期审查指标 |
| 公开来源看不到的潜在执法或诉讼 | 多司法辖区 | 已抓取资料集中未发现重大案件 | 低中 | 中 | 合规栈可能降低风险敞口,但公开沉默不是证据 | 未知,因为可能存在私人纠纷、监管沟通或和解 | 由律师牵头,在关键司法辖区和客户使用场景上检索诉讼和执法记录 |
各行按其可能传导到客户留存和估值的程度排序,而不是按抽象法律新颖性排序。
[CR005, CR006, CR007, CR009, CR010, CR013]监管、端点规则与外部站点控制交汇处,剩余风险最高。
这些序数评分综合 Oxylabs 政策、评论渠道和外部监管 / 平台信号,并非来自内部损失数据。
[CR014, CR018, CR020, CR021, CR024, CR025]7.2 运营、可靠性、安全和供应风险
从运营看,最重要的风险不是工厂或实体设施,而是开放 Web 越来越不友好时,Oxylabs 能否让高成功率数据采集栈持续跑起来。信任中心、Scraper API 和 Web Unblocker 的 SOC 2 Type 2 披露、ISO 27001 认证都是正面信号;DPA 也包含分处理方、数据泄露通知、删除或返还义务,符合企业客户预期。但产品承诺本身把门槛抬高了。Web Unblocker 主打动态策略发现,Web Scraper API 主打结构化、即用型输出;客户一旦依赖这些层,反机器人升级、目标站点改动或基础设施不稳定,都可能很快变成商业痛点。Cloudflare 的按爬取付费和默认拦截调整,就是这种压力的直接例子。公司还依赖合乎伦理来源的住宅代理供给,并依赖第三方网络参与者持续同意且商业上可用。公开来源没有显示 Oxylabs 出现重大泄露或事件,这在方向上令人安心,但没有公开事件证据并不等于运营风险低。剩余敞口因此处于中到高:Oxylabs 看起来比许多代理同行更成熟,但业务仍要靠外部站点、外部网络和持续技术适配来维持客户工作负载运行。[CR002, CR003, CR004, CR011, CR012, CR020]
| 失效模式 | 可能性 | 严重性 | 缓释成熟度 | 残余风险 | 未解决缺口 |
|---|---|---|---|---|---|
| 目标站点封锁、发布方变现,或访问经济性下降 | 高 | 高 | 中 | 重大 | 需要成功率韧性、成本转嫁和目标站点集中度证据 |
| 住宅代理供给退化或来源挑战 | 中 | 高 | 中 | 重大 | 需要参与者来源、流失、地域组合和备用容量细节 |
| 安全事件或子处理方故障 | 中 | 高 | 中高 | 中等 | 需要子处理方清单、事件历史、补丁 / 响应 SLA 和审计证据 |
| ISP、DDoS 或更广泛网络事件导致可用性中断 | 中 | 中高 | 中 | 中等 | 需要正常运行时间历史、事件复盘和客户 SLA / 服务抵扣数据 |
| 支持、退款或入门上的客户预期错配 | 中 | 中 | 中 | 中等 | 需要按队列拆分投诉类别和支持指标 |
Oxylabs 最难的运营问题来自对抗性强且不断变化的网络环境,而不是搭建产品第一版。
[CR002, CR003, CR004, CR008, CR011, CR012]7.3 依赖、客户和市场传导风险
Oxylabs 的风险传导链穿过依赖项和客户预期。公司依赖一整条链:住宅 IP 参与者、目标站点容忍度、云和互联网基础设施、第三方分处理方,以及“Oxylabs 值得溢价付费”的市场叙事。溢价叙事确实存在——Proxyway 和 Gartner 都把 Oxylabs 描述为尤其适合企业或高流量用户——但它并不普遍。Trustpilot、Thunderbit 和低价竞品页面显示,小型或更价格敏感的用户可能会以很不同的方式体验平台,也可能难以忍受 KYC、上手摩擦或高端定价。Webshare 拓宽了渠道触达和低端获客,但也放大了滥用管理和支持复杂度风险。Wiser、Epicup、Environmental Protection Department 等具名客户证明了嵌入式工作流使用,商业上是好事;但这也意味着宕机、法律限制或性能恶化会直接打断客户运营。公开证据在头部客户集中度、净留存率(NRR)和分部收入结构上仍然最弱,因此还无法有信心地建模公司抵御依赖冲击的能力。[CR020, CR021, CR022, CR023, CR024, CR025]
| 依赖 | 交易对手 / 类别 | 作用 | 集中度 | 失效情景 | 严重性 | 缓释措施 | 残余风险 |
|---|---|---|---|---|---|---|---|
| 目标站点访问和平台容忍度 | 发布方、平台、反机器人供应商、受 Cloudflare 保护的站点 | 它们决定能采集多少数据,以及成本多高 | 可能较高但未披露 | 封锁、速率限制或付费访问会降低成功率,并推高服务成本 | 高 | 持续技术适配和高端支持 | 高 |
| 住宅 IP 参与者和获源渠道 | 终端用户 / 获源合作伙伴 | 提供合乎伦理来源的住宅 IP 库 | 未披露 | 参与者流失或获源审查降低库存质量或覆盖面 | 高 | 伦理话术和补偿姿态 | 中高 |
| 云 / 互联网 / 子处理方栈 | 子处理方、ISP、基础设施供应商 | 托管、路由或处理客户工作流的部分环节 | 未披露 | 宕机或控制失效会中断服务,或引发合规事件 | 中高 | 信任中心、DPA 和企业控制 | 中 |
| Webshare 渠道和低端获客路径 | Webshare 品牌 / 自助漏斗 | 扩大对预算敏感用户的覆盖 | 中等 | 更高滥用、支持负荷或低价值流失会削弱单位经济性 | 中 | 品牌隔离和漏斗差异化 | 中 |
| 企业采购信任界面 | 评论平台、认证、文档 | 支撑大买家转化 | 中等 | 任何可见的信任失败、支持走弱趋势或认证问题都会伤害赢单率 | 中高 | SOC 2、ISO 27001、信任中心、KYC 姿态 | 中 |
最关键的依赖来自外部且动态变化:目标站点、IP 供给和信任。
[CR002, CR003, CR004, CR012, CR020, CR021]Oxylabs 依赖几类外部对象,它们可能单独或共同拖累服务质量。
这里按类别而非具名保密供应商展示依赖,因为公开资料无法完整还原子处理方和供应商图谱。
[CR011, CR020, CR021, CR022, CR023, CR033]7.4 人员、财务模型和投资逻辑失效风险
最后一层风险是新预期下的执行。Oxylabs 历史上大多靠自举发展,直到 2026 年 7 月才拿到首笔外部投资;战略上这是好事,但也抬高了运营门槛。机构资本通常要求更紧的报告、更清晰的治理,对法律模糊和支持不一致的容忍度更低。问题在于,公开证据仍留下重大投资核验缺口:没有公开净留存率(NRR)、毛利率、现金消耗、头部客户集中度或分处理方清单,虽然其中部分材料似乎可私下索取。主要财务模型风险因此不是短期资不抵债,而是高端定价、合规成本和平台封锁压力挤压毛利或拖慢增长,速度超过新估值所能承受。投资逻辑会在几种情况下失效:监管者或大型平台实质性压窄可允许的数据采集空间;Oxylabs 的成功率优势明显退化;非企业队列的评价和留存证据恶化;或私下尽调显示经济性明显弱于公开规模故事。正确的投资姿态,是把合规强度视为缓释因素,而不是当前经济证明的替代品。[CR023, CR024, CR025, CR028, CR032, CR038]
| 角色 / 职能 | 依赖或缺口 | 可能性 | 严重性 | 缓释措施 | 尽调路径 |
|---|---|---|---|---|---|
| 高管 / 投资者汇报纪律 | 首次外部融资抬高了治理、预测和董事会沟通门槛 | 中 | 中高 | 规模较大且长期自举经营,可能有助于纪律 | 索取董事会节奏、KPI 包和融资后运营计划 |
| 风险、法律和滥用运营 | AI 和自助需求增长可能超过 KYC 与使用场景审核能力 | 中高 | 高 | 正式 KYC 和 AUP 提供政策抓手 | 索取团队规模、升级处理量、拒绝率和审核 SLA |
| 客户成功和支持 | 企业买家依赖响应迅速的支持,而长尾用户的公开情绪更分化 | 中 | 中高 | 高端定位和正面的企业评论 | 按细分市场索取支持人员配置、CSAT/NPS 和投诉趋势 |
| 跨产品执行 | 高端代理、解锁器、API、AI 工具和 Webshare 渠道扩大了范围,也提高了协调需求 | 中 | 中 | 经营历史长,网络数据品类聚焦 | 要求提供跨品牌产品路线图治理和整合 KPI |
执行风险不在创始团队可信度,而在融资后审视更严,公司能否足够快地把扩张控制补上。
[CR006, CR023, CR024, CR026, CR032, CR033]| 风险 | 可监控触发项 | 阈值 / 事件 | 行动含义 |
|---|---|---|---|
| 隐私 / 抓取监管 | 监管机构正式行动、最终限制性指引,或大客户法律层面的强烈反弹 | EDPB 指引对 Oxylabs 关键工作流明显收紧,或大客户在法律审查期间暂停部署 | 从可跟踪的监管风险升级为威胁投资逻辑的合规风险 |
| 平台封锁 / 发布方变现 | 关键目标类别成功率下降或成本上升 | 持续劣化,或付费访问成本无法转嫁 | 下调估值倍数,增加敞口前要求韧性证据 |
| 客户质量恶化 | 评价结构恶化,或企业客户背书停止扩展 | 退款 / 支持投诉明显增加,且没有新增企业客户证据抵消 | 假设长尾留存更弱、支持负担更高 |
| 安全 / 事件处置失败 | 公开数据泄露、长时间宕机,或认证失效 | 事件坐实,但缺少有力响应证据 | 暂停投资判断,直到弄清事件处理和客户影响 |
| 经济性不透明 | 管理层在尽调中无法提供 NRR、利润率、集中度或现金数据 | 私有数据仍缺失,或明显削弱公开规模叙事 | 认定公开叙事不足以支撑溢价定价 |
| 滥用控制失败 | 因不道德客户用法出现制裁、滥用或重大媒体报道 | 审查薄弱成模式,或有害用例反复出现 | 将合规姿态从缓释因素重评为风险放大器 |
这些触发项用于区分可管理摩擦和会打穿投资逻辑的事件。
[CR006, CR015, CR016, CR020, CR021, CR024]运营和法律冲击很可能先传导到客户留存和利润率,再传导到融资与估值。
该图展示从商业模式推断出的主要因果通道;公开资料未披露实际历史损失树概率。
[CR021, CR022, CR024, CR028, CR036, CR040]7.5 展品
08估值
8.1 投资逻辑与反向逻辑
Oxylabs 的乐观情景可信。它是一家自举公司,在拿外部资本之前就做到约 $350M 年经常性收入(ARR);公司现在声称拥有超过 350,000 名客户或技术团队,客户证明也足够真实,能显示产品已部署在生产工作流,而不是玩具用例。公司所在类别也因 AI 热潮变得更具战略重要性:这类基础设施让模型和智能体持续连接实时公开 Web 数据。规模、相关性,以及在争议行业里偏合规的品牌,合在一起,可以支撑相对增长更慢或更商品化数据供应商的溢价。反向逻辑在于,投资人仍然不了解这种规模的质量。公开材料没有披露净留存率(NRR)、毛利率、现金消耗、服务投入强度或头部客户集中度。就连客户数叙事也有多种定义,包括官网首页的 15,000+ 个客户和融资材料里的 350,000+ 个技术团队。因此,正确的估值问题不是“Oxylabs 是不是好公司?”,而是“当前价格是否已经假设了尚未公开证明的软件级经济性?”[CV001, CV002, CV003, CV004, CV005, CV018]
| 维度 | 乐观逻辑 | 悲观反向逻辑 |
|---|---|---|
| 规模 | ~$350M ARR 以及 350,000+ 客户 / 技术团队叙事,显示品类已有可观规模和相关性 | 公开利润率和留存指标缺失,规模质量仍不清楚 |
| 市场地位 | Oxylabs 卖的是智能体网络基础设施,AI 周期中可能获得溢价估值 | 大平台封锁、隐私压力和法律模糊性可能压住终局倍数 |
| 客户证据 | 具名客户证据和高端评价信号表明企业场景有真实效用 | 长尾评价信号参差,客户数定义也不一致 |
| 经济性 | 十年自举增长暗示资本效率不错,也可能已盈利 | 没有公开 NRR、毛利率或现金消耗数据,经济性可能弱于叙事 |
| 估值 | 10.3x ARR 低于顶级 AI / 基础设施公司;若质量确实高,仍可能留出空间 | 10.3x 仍远高于成熟数据供应商倍数;若质量不及预期,可能压缩 |
| 退出路径 | 大额首轮融资能创造战略可选性,也可能带来未来 M&A 或二级市场兴趣 | 本轮本身可能就是稀缺溢价,留给后续投资者的上行空间更少 |
投资逻辑不只是 Oxylabs 是否重要,而是它是否配得上公开软件公司级别的经济性。
[CV001, CV003, CV005, CV012, CV018, CV019]建议链条先看业务质量,再看不透明度,最后落到价格纪律。
该流程呈现由公开证据推导出的决策逻辑,不是机械评分模型。
[CV005, CV018, CV020, CV023, CV029, CV041]8.2 融资背景和公开可比区间
融资背景表面上异常干净:Oxylabs 首次外部融资拿到 $130M,估值 $3.6B;此前公司称已建成超过 $350M 的 ARR。单看这一点,约等于 10.3x ARR 倍数。问题不在于这个倍数无法辩护,而在于公开可比公司的区间太宽,正确答案高度依赖隐藏的质量指标。Similarweb 约 2.3x 销售额、ZoomInfo 约 0.7x,说明即使数据供应商仍有可观规模,只要增长更慢或挑战更多,公开市场也会打折。C3.ai 约 4.6x,高于成熟信息供应商,但仍远低于高端基础设施龙头。另一端,Cloudflare 和 Palantir 约 45.7x 和 61.7x,反映的是围绕品类领导力、毛利率、可持续 AI 或基础设施溢价的强得多的公开叙事。因此,Oxylabs 落在一个尴尬但分析上有用的中间地带:明显贵过成熟数据供应商,明显便宜过顶级 AI 或连接性公司,同时仍足够不透明,投资人还无法判断它到底应落在区间里的哪一段。[CV001, CV002, CV003, CV004, CV005, CV006]
| 可比公司 | 指标 | 倍数 / 估值 / 状态 | 相关性 | 局限 |
|---|---|---|---|---|
| Similarweb | 2026 P/S 比率和 Q1 经营画像 | ~2.28x P/S;Q1 2026 收入增长 10%,非 GAAP 经营利润为正 | 数字情报和网络数据邻近性上最接近的公开可比公司 | 网站分析和市场情报组合更宽,较少以代理基础设施为中心 |
| ZoomInfo | 2026 P/S 比率和 Q1 利润率画像 | ~0.73x P/S;Q1 2026 收入增长 1.5%,调整后经营利润率 35% | 说明公开市场会如何大幅折价增长较慢的大型数据供应商 | 市场进入情报更偏销售科技,而不是网络数据基础设施 |
| C3.ai | 2026 P/S 比率和 FY2026 业绩 | ~4.60x P/S;FY2026 收入 $250.3M,聚焦重组和降本 | 可作为中档 AI 软件参照,适合仍在证明经济性持久的公司 | 应用软件模式与 Oxylabs 的数据获取栈差异明显 |
| Cloudflare | 2026 P/S 比率和 Q1 经济性 | ~45.7x P/S;Q1 2026 收入 $639.8M,增长 34%,毛利率 71.2% | 代表有 AI 顺风的高端互联网基础设施估值上沿 | 平台更宽、公开披露更强,规模验证也强于 Oxylabs |
| Palantir | 2026 P/S 比率和 Q1 监管文件背景 | ~61.7x P/S;异常强的毛利率和 AI 叙事支撑溢价 | 展示顶级数据 / AI 平台的溢价天花板 | 业务组合、政府敞口和披露深度都与 Oxylabs 相距很远 |
可比公司组刻意设得较宽,因为没有一家上市公司能与高端代理加网络数据基础设施完全一一对应。
[CV006, CV007, CV008, CV009, CV010, CV011]公允价值高度取决于投资者愿意给大约 $350M ARR 套用多少倍数。
数值是以十亿美元计的企业价值口径简写,按简单收入倍数套用约 $350M ARR 锚点。
[CV005, CV025, CV026, CV027, CV028]8.3 情景估值与建议
我们的基准情景从一个足够清楚的公开锚点出发:2026 年 7 月融资时约 $350M ARR。如果 Oxylabs 真是一家资本效率高、企业级的基础设施公司,合规控制可信、留存强,那么相对 Similarweb、ZoomInfo 甚至 C3.ai 的溢价都说得通。但在毛利率、流失率和集中度可见之前,很难给出接近 Cloudflare 或 Palantir 的公有云式倍数。由此得到一个务实的情景框架。悲观情景假设公司更像一家位置不错但有法律敞口的数据供应商,价值可能被压到约 5x-7x ARR。基准情景假设经济性有溢价但未到顶级,支撑约 8.5x-11.5x ARR。乐观情景假设管理层能证明软件式留存、健康毛利和可持续 AI 顺风,支撑约 12x-16x ARR。按这个基础,2026 年 7 月 $3.6B 估值落在基准情景区间内。因此,当前证据支持“观察”建议,信心中等、风险中高、估值立场合理;不是只因 AI 叙事令人兴奋就给出买入。[CV005, CV012, CV018, CV019, CV023, CV025]
| 维度 | 评估 | 依据 |
|---|---|---|
| 建议 | 观察 | 业务质量清楚,但仅凭公开证据,当前估值并非明显错定价 |
| 信心 | 中 | 估值依赖已披露 ARR、外部轮价格,以及不完美的公开可比公司替代指标 |
| 风险评级 | 中高 | 监管、平台和经济透明度风险仍然重要 |
| 估值立场 | 合理 | $3.6B 估值落在基于公开证据推导的基准情景区间内 |
| 决策含义 | 只有深入私有尽调后才推进 | 买入逻辑需要留存、利润率、现金生成和集中度证据 |
| 价格敏感度 | 高 | 若入场价明显更低,或私有指标更强,结论可能很快改善 |
建议对价格和证据都敏感,不是笼统的业务质量打分。
[CV028, CV029, CV041]| 情景 | 假设 | 估值 / 回报逻辑 | 关键风险 | 概率信号 |
|---|---|---|---|---|
| 乐观 | ARR 质量高;NRR 和毛利率证明其有类软件经济性;AI 需求继续加速;合规姿态支撑高端企业采用 | 12x-16x ARR 乘以 ~$350M,对应约 $4.2B-$5.6B | 当前不透明可能掩盖较弱经济性;公开可比公司溢价可能均值回归 | 需要留存、利润率和集中度的强私有证据 |
| 基准 | 业务质量真实,但并非完全顶级;法律和平台风险仍可控;利润率足够好,但达不到公有云那样的超高水平 | 8.5x-11.5x ARR 对应约 $3.0B-$4.0B | 若 ARR 质量被夸大,投资者仍可能付高价 | 最符合当前公开证据 |
| 悲观 | 增长放缓,法律 / 平台摩擦推高服务成本,或客户持久性弱于暗示 | 5x-7x ARR 对应约 $1.8B-$2.5B | 经济性透明后,当前估值被证明过于慷慨 | 留存弱、集中度高,或倍数向较低公开可比公司压缩时触发 |
情景区间使用 ARR 倍数,因为公开盈利数据过于不完整,无法支撑更精确的 DCF 或 rule-of-40 框架。
[CV005, CV025, CV026, CV027, CV028, CV030]当前轮次最好理解为处在基准情景区间内;如果质量指标不及预期,下行空间不小。
区间以十亿美元计,由公开 ARR 和各情景倍数假设推导。
[CV025, CV026, CV027, CV028]公司在市场相关性和客户证据上得分较高,但经济性透明度得分很差。
分数是基于公开证据集作出的分析师判断,用来组织讨论,不代表统计精确性。
[CV018, CV019, CV020, CV023, CV029, CV041]8.4 最终尽调和投资逻辑失效触发点
作出明确买入建议之前,剩余工作集中在经济真实性核验。投资人需要截至 runDate 的经审计 ARR 桥、按产品和细分切分的留存视图、毛利率拆解、客户集中度分析、现金与烧钱速度、服务或支持强度的角色,以及 Warburg 新一轮融资带来的精确优先权或悬置影响。这些问题重要,是因为 10.3x ARR 估值既不明显便宜,也不明显鲁莽;只有缺失的质量指标足够强,它才有吸引力。投资逻辑会在几种情况下失效:ARR 数字包含过多非经常性或间接量;大客户高度集中;法律或平台摩擦显著抬高服务成本;或高端队列之外的客户满意度脆弱到无法支撑持久扩张。如果下一轮融资或流动性事件必须以低得多的倍数成交,因为公开市场可比公司向区间低端收敛,逻辑也会失效。在这些问题回答之前,纪律比赞赏更重要。[CV021, CV022, CV023, CV024, CV031, CV032]
| 触发项 | 阈值 | 对投资逻辑的传导 | 行动含义 |
|---|---|---|---|
| ARR 质量不及预期 | ARR 桥接显示一次性、低质量或非经常性成分较大 | 直接削弱 10.3x ARR 锚点 | 除非价格大幅重置,否则从观察转为放弃 |
| 留存或集中度走弱 | NRR、流失率或头部客户敞口明显差于预期 | 推动估值向公开可比区间低端靠拢 | 用悲观情景倍数重新承销 |
| 法律或平台摩擦上升 | 平台封锁或监管约束明显推高服务成本 | 削弱利润率质量和长期可扩展性 | 下调公允价值区间,并要求缓释证据 |
| 客户体验恶化 | 企业客户证据停滞,同时退款 / 支持投诉更显眼 | 暗示旗舰客户群之外的持久性更弱 | 假设扩张更温和、溢价更低 |
| 后续融资以更低价格成交 | 下一轮或二级交易在缺少强外生解释时低于当前估值成交 | 暗示当前价格更多包含稀缺性,而非基本面 | 将本轮视为私募估值峰值 |
| 治理或优先权悬空问题恶化 | 轮次条款或未来优先股堆叠明显稀释普通股等效上行 | 即便企业价值守住,也会压低投资者回报 | 收紧入场纪律,或回避 |
这些触发项应在尽调或后续融资更新中可观察,而不是笼统的定性担忧。
[CV005, CV021, CV023, CV031, CV039, CV040]| 主题 | 缺失证据 | 重要性 | 负责人或尽调路径 |
|---|---|---|---|
| ARR 桥接 | 从账单、订单额和收入到所称 ARR 的审计后勾稽 | 当前估值比任何其他变量都更依赖 ARR | 与 CFO 核验财务,并审查经审计收入明细 |
| 留存队列 | 按产品和分部拆分的 NRR、GRR、Logo 流失和队列扩张 | 决定 Oxylabs 配得上溢价,还是只配稳健的私募倍数 | 客户分析和董事会 KPI 包 |
| 毛利率和服务成本 | 产品级、渠道级毛利率,以及基础设施成本驱动因素 | 说明业务更像软件经济性,还是运营更重的基础设施 | 财务与工程尽调 |
| 客户集中度 | 按 ARR、用量和续约日期拆分的 Top-10、Top-20 敞口 | 客户基数很大,仍可能藏着企业客户集中风险 | 收入运营和财务审查 |
| 融资轮经济条款 | 优先权条款、老股 / 新股拆分、治理权利和稀释压力 | 单看企业价值无法决定投资者回报 | 法务和股权结构表尽调 |
| 监管 / 平台韧性 | 对目标站点集中度、封锁率、KYC 执行和合规例外处理的内部判断 | 一旦边界摩擦出现,估值下行会由它驱动 | 产品、法务和风险运营尽调 |
这些事项是把合理的公开叙事转成可承销投资判断的最短路径。
[CV023, CV024, CV032, CV038, CV039, CV041]8.5 展品
免责声明
本报告仅供参考,反映截至 2026-07-19 可获得的公开资料,不构成投资建议。私营公司估值、ARR 声明和情景区间应在任何投资决定前独立核验。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Oxylabs was established in 2015. | 高 | SO002, SO003, SO010 |
| CO002 | Oxylabs is headquartered in Vilnius, Lithuania. | 高 | SO003, SO010, SO029 |
| CO003 | Oxylabs says it was founded within the Tesonet ecosystem. | 中 | SO010 |
| CO004 | Oxylabs has evolved from a premium proxy provider into a broader web-data infrastructure platform. | 高 | SO002, SO003, SO005 |
| CO005 | Oxylabs positions its platform as foundational infrastructure for AI, agentic applications, and enterprise web intelligence. | 高 | SO003, SO005 |
| CO006 | Vytautas Savickas is the current Chief Executive Officer of Oxylabs. | 高 | SO003, SO011 |
| CO007 | Founder Julius Černiauskas transitioned from CEO to Chairman of the Board in 2026. | 中 | SO011 |
| CO008 | Jurgis Gabrielius Rudgalvis is publicly identified as Oxylabs’ Chief Financial Officer. | 高 | SO003, SO010 |
| CO009 | On 2026-07-09 Oxylabs announced a $130 million investment from Warburg Pincus. | 高 | SO002, SO003, SO005 |
| CO010 | The Warburg Pincus investment valued Oxylabs at approximately $3.6 billion. | 高 | SO002, SO003, SO005, SO006 |
| CO011 | The 2026 investment was made through the Warburg Pincus Capital Solutions Founders Fund. | 高 | SO003, SO005 |
| CO012 | Oxylabs was bootstrapped from founding until its July 2026 financing round. | 高 | SO003, SO005, SO007, SO010 |
| CO013 | Oxylabs says it is recording $350 million in ARR. | 高 | SO002, SO010, SO005 |
| CO014 | Warburg and Tesonet coverage say Oxylabs has over 350,000 customers. | 高 | SO003, SO010, SO005 |
| CO015 | The Oxylabs homepage says the company is trusted by 15,000+ clients globally. | 中 | SO001 |
| CO016 | The Oxylabs compliance page says 500+ experts work at the company. | 中 | SO013 |
| CO017 | Public sources do not clearly define how Oxylabs distinguishes 350,000+ customers from 15,000+ clients. | 中 | SO001, SO003, SO010, SO013 |
| CO018 | Oxylabs’ current platform includes proxies, scraping APIs, data products, and AI-facing tooling rather than a single proxy product. | 高 | SO001, SO019, SO020 |
| CO019 | The Residential Proxies product page advertises 175M+ IPs across 195 countries with plans starting at $6 per GB. | 中 | SO016 |
| CO020 | The Datacenter Proxies page advertises 1.9M+ dedicated IPs, 188-country coverage, and 99.9% success and uptime claims. | 中 | SO018 |
| CO021 | The Mobile Proxies page advertises 20M+ mobile IPs in 140+ countries. | 中 | SO017 |
| CO022 | The Web Scraper API page shows Oxylabs selling structured web-data extraction as a managed product starting at $49 per month and integrated with AI platforms. | 高 | SO019, SO020 |
| CO023 | Oxylabs’ 2025 impact report says the company acquired ScrapingBee, launched AI Studio, introduced the Trust Center, and expanded its patent portfolio to 150 in 2025. | 中 | SO012 |
| CO024 | By July 2026 Oxylabs said its patent portfolio exceeded 160 patents. | 高 | SO002, SO003 |
| CO025 | Webshare announced that Oxylabs had completed its acquisition of Webshare Software in 2022. | 中 | SO021 |
| CO026 | Oxylabs publicly presents compliance, KYC, ethics, and proxy-source legitimacy as core parts of its operating model. | 高 | SO002, SO013, SO015 |
| CO027 | Oxylabs publicly references ISO 27001 and SOC 2 security and compliance credentials on its trust and risk materials. | 高 | SO013, SO014 |
| CO028 | The 2025 impact report says Project 4β supported organizations including Bellingcat and the Pulitzer Center in 2025. | 中 | SO012 |
| CO029 | The Oxylabs homepage shows active July 2026 product-update cadence including agent skills, headless-browser improvements, and batch requests for LLM targets. | 中 | SO001 |
| CO030 | Oxylabs’ AI Predictions 2026 content frames agentic browsing, governance, and reliable public-web access as major themes for the next AI wave. | 中 | SO026 |
| CO031 | Oxylabs used DEVWorld 2026 and the AI Engineer World’s Fair to market its role in AI and developer web-data workflows. | 中 | SO027, SO028 |
| CO032 | Trustpilot reviews include complaints about refunds, support quality, and account blocks tied to KYC or security checks. | 中 | SO025 |
| CO033 | G2 reviews rate Oxylabs 4.5 out of 5 from 421 reviews while also noting pricing pressure for smaller users. | 中 | SO023 |
| CO034 | Capterra reviews praise reliability and support but repeatedly cite premium pricing and some feature gaps for smaller users. | 中 | SO024 |
| CO035 | TNW says Oxylabs handles billions of requests a day. | 中 | SO005 |
| CO036 | Warburg and Oxylabs describe the company as serving blue-chip or Fortune 500 customers. | 高 | SO003, SO005 |
| CO037 | Oxylabs said the Warburg financing would expand its capacity for future acquisitions and partnerships. | 中 | SO002 |
| CO038 | Oxylabs’ CFO said Webshare and ScrapingBee were already success stories for product-portfolio enhancement. | 中 | SO002 |
| CO039 | Oxylabs markets use cases including AI training and RAG, ad verification, cybersecurity, travel fare aggregation, and e-commerce intelligence. | 高 | SO001, SO016, SO017 |
| CO040 | Oxylabs publicly names governance, risk, privacy, litigation, and procurement personnel on its risk and legal page. | 中 | SO013 |
| CO041 | Craft currently lists Julius Černiauskas as Oxylabs’ Chief Executive Officer. | 低 | SO029 |
| CO042 | Craft currently lists Oxylabs as founded in 2012. | 低 | SO029 |
| CO043 | The discrepancy between Craft and current Oxylabs disclosures suggests some third-party data providers lag real-time company changes. | 中 | SO011, SO029 |
| CO044 | Tesonet says Oxylabs became the second bootstrapped unicorn in its ecosystem and Lithuania’s sixth unicorn overall. | 中 | SO010 |
| CO045 | Tech Funding News and TNW frame Oxylabs as Lithuania’s second most valuable private company after Vinted or one of the country’s most valuable private companies. | 高 | SO005, SO006 |
| CO046 | Public transaction disclosures name Goldman Sachs as sole placement agent to Oxylabs and Deloitte Lithuania as financial and tax advisor. | 高 | SO003, SO010 |
| CO047 | Public transaction disclosures name Tegos, Norton Rose Fulbright, Lazard, Latham & Watkins, Freshfields, and Sorainen among the legal and financial advisors on the 2026 deal. | 高 | SO003, SO010 |
| CM001 | For Oxylabs, the relevant market includes web-scraping software, proxy access, anti-bot bypass, browser automation, and adjacent public-web data infrastructure rather than proxy resale alone. | 中 | SM004, SM009, SM013 |
| CM002 | The Business Research Company sizes the global web scraping market at $1.17 billion in 2026 and $2.28 billion by 2030. | 中 | SM004 |
| CM003 | Mordor Intelligence sizes the global web scraping market at $1.56 billion in 2026 and $3.49 billion by 2031. | 中 | SM003 |
| CM004 | The gap between the TBRC and Mordor estimates shows that market-size outputs vary materially with market-definition choices. | 中 | SM003, SM004, SM007 |
| CM005 | TBRC defines the web scraping market as software and services sold across cloud and on-premise deployment models. | 中 | SM004 |
| CM006 | TBRC says both large enterprises and SMBs buy web scraping solutions. | 中 | SM004 |
| CM007 | TBRC lists price monitoring, market research, data aggregation, lead generation, sentiment analysis, and competitive intelligence as core web-scraping applications. | 中 | SM004 |
| CM008 | TBRC names retail and e-commerce, BFSI, media, travel, manufacturing, IT, healthcare, real estate, and government as web-scraping end-user verticals. | 中 | SM004 |
| CM009 | TBRC says North America was the largest web-scraping region in 2025 and Asia-Pacific is the fastest-growing region. | 中 | SM004 |
| CM010 | TBRC names Oxylabs, Bright Data, Zyte, Apify, ScraperAPI, ScrapingBee, Smartproxy, and IPRoyal among major companies in the web scraping market. | 中 | SM004 |
| CM011 | TBRC identifies AI-powered low-code scraping tools as a major market trend and uses Oxylabs AI Studio as an example. | 中 | SM004 |
| CM012 | In Apify’s 2026 survey, 65.8% of respondents reported increased proxy usage and 58.3% reported higher proxy spending year over year. | 中 | SM006 |
| CM013 | More than 62% of Apify survey respondents reported increased infrastructure spending, largely because of stronger anti-bot protections. | 中 | SM006 |
| CM014 | Apify’s 2026 survey says 45.8% of respondents already use AI in scraping workflows while 54.2% do not. | 中 | SM006 |
| CM015 | Among Apify respondents already using AI for scraping, 72.7% report productivity benefits and all say they plan to increase usage. | 中 | SM006 |
| CM016 | Apify’s survey says 66.2% of respondents plan to try AI-assisted scraping tools in the future. | 中 | SM006 |
| CM017 | ScrapingDog summarizes 2026 market context as a multi-billion-dollar web-scraping sector with a $4.9 billion alternative-data adjacency growing 28% annually. | 低 | SM007 |
| CM018 | ScrapingDog says 39.1% of practitioners use proxy providers, 34.8% use scraping APIs, and 26.1% use cloud-based scraping platforms. | 低 | SM007 |
| CM019 | ScrapingDog says Python accounts for 69.6% of scraping-language usage, with BeautifulSoup, Crawlee, Selenium, and Playwright among the common tools. | 低 | SM007 |
| CM020 | Cybernews’ summary of Proxyway’s 2026 report says some major proxy providers are growing 50% or more year over year because of AI demand. | 中 | SM005 |
| CM021 | Cybernews says AI has become the largest customer vertical by volume for some major proxy providers. | 中 | SM005 |
| CM022 | Cybernews says more than 50 proxy-server vendors were established in 2025, indicating low entry barriers. | 中 | SM005 |
| CM023 | Cybernews says access to residential proxies still costs up to 75% less than three years earlier, showing ongoing commoditization. | 中 | SM005 |
| CM024 | Cybernews says established providers increasingly compete by serving enterprise users and expanding vertically into less commoditized scraping tools. | 中 | SM005 |
| CM025 | Cybernews says botnets and botnet-backed proxy providers have destabilized the market and increased pressure for responsible sourcing. | 中 | SM005 |
| CM026 | OpenWebSearch says 14 European research and computing centers joined to build an open European web-search infrastructure for digital sovereignty. | 中 | SM009 |
| CM027 | OpenWebSearch reports 9.14 billion URLs crawled, 185 languages, 28 million hosts, and roughly 1 TiB crawled per day. | 中 | SM009 |
| CM028 | The Open Web Index beta stores roughly 1 petabyte of open web data and makes it available for search applications and LLMs under a research license. | 高 | SM010, SM026 |
| CM029 | OpenWebSearch frames dependence on a few large search gatekeepers as a risk to unbiased access and European innovation. | 中 | SM009 |
| CM030 | Cloudflare says many publishers want a third path between blocking AI crawlers and allowing them free access: charging for crawl access. | 高 | SM013, SM014 |
| CM031 | Cloudflare’s pay-per-crawl system is in private beta and uses HTTP 402 Payment Required responses to signal priced content access. | 高 | SM013, SM014, SM015 |
| CM032 | DataDome says web scraping costs businesses billions annually by enabling price undercutting, content theft, and competitive intelligence abuse. | 中 | SM019 |
| CM033 | DataDome says e-commerce businesses lose 2% of online revenue annually to web scraping. | 中 | SM019 |
| CM034 | DataDome says only about 7% of tested websites blocked advanced anti-fingerprinting bots in its 2025 bot-security report. | 中 | SM019 |
| CM035 | DataDome says it blocked an 80-million-request scraping attack using 855,000 unique IPs against a review platform in March 2026. | 中 | SM019 |
| CM036 | DataDome says 80% of AI agents do not properly identify themselves when visiting websites and 79.7% of tested sites allowed a spoofed ChatGPT-style user agent through. | 中 | SM020 |
| CM037 | DataDome says nearly 8 billion AI-agent requests hit its network in the first two months of 2026. | 中 | SM020 |
| CM038 | The European Commission’s AI Act summary says high-risk AI systems require high-quality datasets, logging, documentation, human oversight, robustness, and cybersecurity. | 高 | SM018, SM027 |
| CM039 | The AI Act prohibits untargeted internet scraping to create or expand facial-recognition databases. | 高 | SM018, SM027 |
| CM040 | The AI Act’s transparency rules for AI-generated content come into effect in August 2026, while GPAI obligations became effective in August 2025. | 中 | SM027 |
| CM041 | Oxylabs says its compliance program aligns with GDPR, the Digital Services Act, the Cyber Resilience Act, and other European laws, indicating that legal readiness is part of category competition. | 中 | SM002 |
| CM042 | Proxyway’s 2026 residential tests rank Oxylabs as the best premium residential provider with 99.93% success rate and 0.60-second response time, while Decodo is the best value provider. | 中 | SM021 |
| CM043 | AIMultiple says Bright Data and Oxylabs are among the fastest proxy providers in 2026 benchmark testing. | 中 | SM023, SM024 |
| CM044 | AIMultiple says anti-bot sophistication has shifted the proxy market focus from raw IP volume to IP reputation. | 中 | SM023 |
| CM045 | AIMultiple says residential proxy prices average $3 to $15 per GB in 2026, while datacenter proxies can start at $0.5 per IP. | 中 | SM025 |
| CM046 | AIMultiple says Oxylabs charges a premium at entry-level volumes because buyers are paying for larger pools and higher success rates. | 中 | SM025 |
| CM047 | Public APIs, licensed datasets, manual research, in-house stacks, and open web indexes all function as substitutes or complements to full-stack scraping vendors. | 中 | SM009, SM010, SM013, SM025 |
| CM048 | TBRC segmentation and Apify’s survey composition imply that budget ownership spans marketing, research, product, engineering, security, and founder-led SMB use cases. | 中 | SM004, SM006 |
| CM049 | The market shows a two-speed structure: proxy-layer commoditization at the low end and premiumization around managed tooling, compliance, and performance at the high end. | 中 | SM005, SM013, SM019, SM025 |
| CM050 | The AI-agent wave increases demand for real-time public-web infrastructure rather than eliminating the need for it. | 中 | SM001, SM009, SM013, SM020 |
| CP001 | Oxylabs’ competitor set in 2026 includes direct premium peers, value challengers, managed scraping APIs, automation platforms, and internal-build substitutes rather than proxy vendors alone. | 中 | SP018, SP019, SP023, SP025 |
| CP002 | Oxylabs’ homepage says the company is trusted by more than 15,000 clients globally. | 中 | SP001 |
| CP003 | Bright Data and Oxylabs are the most consistently paired premium head-to-head vendors in retained 2026 comparison content. | 中 | SP004, SP013, SP019 |
| CP004 | Bright Data’s 2026 comparison page says Bright Data has 400M+ residential IPs while Oxylabs has 175M+ residential IPs and a 2M+ datacenter pool. | 中 | SP004 |
| CP005 | Bright Data’s own residential proxy page advertises over 400M monthly ethical residential IPs from 195 countries. | 中 | SP007 |
| CP006 | Bright Data says it protects data for more than 20,000 organizations and markets ISO 27001, SOC 2, SOC 3, and CSA STAR credentials. | 中 | SP007 |
| CP007 | Bright Data markets pay-as-you-go purchasing and free trials on key proxy products. | 中 | SP004, SP007 |
| CP008 | Bright Data’s 2026 web-unblocker guide puts Bright Data Web Unlocker at a 97.9% success rate in its cited benchmark framing. | 中 | SP006 |
| CP009 | The same Bright Data guide presents Oxylabs Web Unblocker as enterprise-grade and cites a public starting price around $9.4 per GB. | 中 | SP006 |
| CP010 | Decodo advertises 115M+ ethically sourced residential IPs across 195+ locations. | 中 | SP009, SP010 |
| CP011 | Decodo markets a 99.92% success rate, pay-as-you-go buying, and a short free-trial path. | 中 | SP009, SP010 |
| CP012 | Proxyway’s 2026 residential ranking calls Decodo the best value provider and Oxylabs the best premium residential proxy provider. | 中 | SP018 |
| CP013 | SOAX positions itself around routing quality and predictable pricing rather than pure lowest-cost access. | 中 | SP011, SP012 |
| CP014 | PrivateProxyGuide positions SOAX as a flexible alternative and frames Bright Data as potentially too expensive or complex for narrower needs. | 低 | SP023 |
| CP015 | Apify’s market role is adjacent to proxy infrastructure because its main platform centers a large AI-tool and actor marketplace rather than only traffic resale. | 中 | SP014, SP025 |
| CP016 | Apify’s June 2026 comparison says Bright Data offers 600+ ready-made scrapers and a dataset marketplace, while Oxylabs emphasizes a general scraper API and API playground instead. | 中 | SP013 |
| CP017 | Webshare attacks the entry-level buyer with a permanent free tier of 10 datacenter proxies and simple low-cost proxy packaging. | 中 | SP015, SP022 |
| CP018 | IPRoyal emphasizes pay-as-you-go residential proxy usage and explicitly frames residential proxies as useful for large-scale public-text data collection workflows. | 中 | SP017 |
| CP019 | Rayobyte competes more aggressively on low-cost datacenter and ISP economics plus easy trials than on the broad premium workflow stack shown by Oxylabs or Bright Data. | 中 | SP020 |
| CP020 | ScraperAPI offers a free 1,000-credit monthly plan and premium proxy pools, making it a managed-API substitute for buyers who do not want to procure raw proxy traffic directly. | 中 | SP016 |
| CP021 | ScrapingBee’s 2026 guide argues that many teams prefer managed rotation, rendering, and anti-bot handling over fine-grained control of raw proxies. | 中 | SP021 |
| CP022 | AIMultiple says the market focus is shifting from raw IP volume to IP reputation, tested performance, and price-efficiency. | 中 | SP019 |
| CP023 | BestProxyCompare’s 2026 ranking putting a low-cost vendor at number one shows that pure price-performance buyers can bypass premium incumbents entirely. | 低 | SP024 |
| CP024 | Independent rankings consistently place Oxylabs in the premium provider bucket rather than the best-value bucket. | 中 | SP018, SP019, SP023 |
| CP025 | Bright Data and Oxylabs both cover the core proxy stack plus scraper tooling, but Bright Data’s public product surface extends further into datasets, browser tooling, and agent workflows. | 中 | SP004, SP013 |
| CP026 | Bright Data’s dataset marketplace, web IDE, and agent-oriented tools create differentiation above raw proxy resale. | 中 | SP004, SP013 |
| CP027 | Oxylabs narrows the platform gap with AI Studio, Web Scraper APIs, and Web Unblocker, but retained public evidence still shows less marketplace-style breadth than Bright Data. | 中 | SP001, SP003, SP004, SP013 |
| CP028 | Standard proxy protocols, usage-based packaging, and trial mechanics make multi-homing practical for many proxy buyers. | 中 | SP010, SP016, SP017, SP022 |
| CP029 | Switching costs rise when a buyer depends on unblockers, tuned anti-bot logic, support teams, or workflow-specific integrations rather than basic traffic routing alone. | 中 | SP003, SP006, SP021 |
| CP030 | Publicly visible distribution power differs by vendor class: Bright Data leans on enterprise breadth and compliance proof, Apify on marketplace distribution, and Oxylabs on premium enterprise positioning. | 中 | SP001, SP007, SP014 |
| CP031 | Trust and regulatory posture are explicit buying criteria across the premium market, with Bright Data advertising certifications and Decodo and Oxylabs emphasizing ethical sourcing and compliance language. | 中 | SP001, SP007, SP009 |
| CP032 | Distill’s competitor graph grouping Apify, Oxylabs, and SOAX around Bright Data supports a landscape broader than one-to-one proxy rivalry. | 低 | SP025 |
| CP033 | PrivateProxyGuide explicitly argues that buyers may leave Bright Data for cheaper, simpler, or more specialized alternatives. | 低 | SP023 |
| CP034 | Competitor-authored comparison pages portray Oxylabs as premium and capable, but not as the cheapest or most accessible option for small buyers. | 中 | SP004, SP013 |
| CP035 | Webshare, IPRoyal, and similar entry-level offers pressure Oxylabs at the low end on price and onboarding simplicity. | 中 | SP015, SP017, SP022, SP024 |
| CP036 | Apify, ScraperAPI, and ScrapingBee can displace proxy-first vendors by selling workflow simplicity or data-collection outcomes instead of infrastructure procurement. | 中 | SP013, SP016, SP021 |
| CP037 | Public evidence does not establish a uniquely proprietary supply-side moat for any leading vendor because network overlap, sourcing concentration, and exclusivity are not disclosed clearly. | 中 | SP004, SP019, SP024 |
| CP038 | Oxylabs’ moat in public evidence appears strongest in premium performance, support, and trust posture rather than in hard lock-in or uniquely exclusive distribution. | 中 | SP001, SP007, SP018, SP019 |
| CP039 | Competitive durability is threatened simultaneously by commodity-price entrants below the market and managed scraping platforms above the proxy layer. | 中 | SP016, SP021, SP023, SP024 |
| CP040 | Public sources are still insufficient to model competitor win rates, network-quality overlap, or customer switching behavior with high precision. | 中 | SP019, SP023, SP024 |
| CI001 | Oxylabs’ July 2026 investment announcement says the company is recording $350 million in ARR and is trusted by more than 350,000 tech teams worldwide. | 中 | SI001 |
| CI002 | Warburg Pincus says the July 2026 transaction marked Oxylabs’ first outside investment since the company’s 2015 founding. | 中 | SI002 |
| CI003 | The use of Warburg Pincus’ Capital Solutions Founders Fund suggests a founder-oriented growth-capital structure rather than a rescue financing or conventional multi-investor venture round. | 中 | SI002, SI008, SI015 |
| CI004 | Warburg and Oxylabs say the investment values the company at approximately $3.6 billion. | 高 | SI001, SI002, SI008 |
| CI005 | Official July 2026 funding materials say Oxylabs plans to use the new capital to extend competitive advantage, accelerate next-generation products and capabilities, and support acquisitions or partnerships. | 中 | SI001, SI002, SI004, SI008 |
| CI006 | TNW, EU-Startups, Tesonet, and IBTimes each repeat the core topline narrative that Oxylabs reached about $350 million ARR and more than 350,000 customers or tech teams by July 2026. | 中 | SI003, SI004, SI005, SI025 |
| CI007 | EU-Startups expresses the same ARR figure as €305.8 million ($350 million), indicating a currency-converted rather than contradictory scale disclosure. | 中 | SI004 |
| CI008 | Oxylabs’ homepage separately says the company is trusted by 15,000+ clients globally. | 中 | SI007 |
| CI009 | The coexistence of 15,000+ clients and 350,000+ tech teams or customers makes public unit-economics inference possible but unreliable without management definitions. | 中 | SI001, SI006, SI007 |
| CI010 | Oxylabs publicly prices residential proxies on a per-GB basis and offers monthly plans with meaningful volume discounts from about $6/GB toward $2.50/GB. | 中 | SI009 |
| CI011 | Oxylabs publicly prices datacenter proxies across both per-IP and per-GB formats, including list points such as $1.20/IP, $0.70/IP, and $0.44/GB. | 中 | SI010 |
| CI012 | Oxylabs publicly prices Web Unblocker at roughly $9.40/GB, $8.60/GB, and $7.50/GB across visible plans. | 中 | SI013 |
| CI013 | Oxylabs’ public Web Scraper API page shows mixed monetization logic that can include per-GB billing and target-specific per-1,000-result pricing such as $0.25 for Amazon and $0.50 for Google. | 中 | SI011 |
| CI014 | Free-trial calls to action appear across Oxylabs residential proxies, datacenter proxies, Web Scraper API, and Web Unblocker pages. | 中 | SI009, SI010, SI011, SI013 |
| CI015 | Larger public plans also advertise dedicated account managers, tailored solutions, and 24/7 support, indicating an enterprise-assisted sales motion rather than purely self-serve monetization. | 中 | SI009, SI011, SI013 |
| CI016 | The public pricing surface shows Oxylabs is a hybrid recurring-and-usage business rather than a classic seat-only SaaS company. | 中 | SI009, SI010, SI011, SI013 |
| CI017 | Proxyway’s search API report shows Oxylabs also monetizes search-oriented API workloads, expanding the revenue model beyond proxy traffic alone. | 中 | SI017 |
| CI018 | The revenue architecture spans raw connectivity, managed anti-bot access, structured data collection, and enterprise support layers. | 中 | SI011, SI012, SI013 |
| CI019 | Webshare’s acquisition preserved a low-price, self-serve motion after the deal, suggesting Oxylabs can serve a budget-oriented funnel without fully diluting the premium Oxylabs brand. | 中 | SI014 |
| CI020 | The 2025 impact report says Oxylabs launched AI Studio and acquired ScrapingBee during 2025, showing continuing investment in higher-layer products. | 中 | SI006 |
| CI021 | Official and partner sources frame the 2026 Warburg round as growth capital for product expansion and strategic deals rather than emergency financing. | 中 | SI001, SI002, SI005 |
| CI022 | Proxyway’s Oxylabs review says residential proxies are sourced through Honeygain, direct ISP partnerships, and app developers. | 中 | SI016 |
| CI023 | Proxyway’s 2026 market-research profile says Oxylabs sources IPs through partner apps Honeygain, JumpTask, and an SDK and competes in the enterprise segment. | 中 | SI020 |
| CI024 | The same Proxyway market-research source estimates Oxylabs at roughly 400-500 employees and lists scraping APIs, browser products, and datasets alongside core proxy networks. | 低 | SI020 |
| CI025 | Public evidence implies the main cost buckets are IP sourcing, cloud or network operations, anti-bot engineering, support, account management, and compliance rather than heavy fixed-asset capex. | 中 | SI006, SI016, SI020 |
| CI026 | Oxylabs’ product pages explicitly market high success rates, low response times, and 24/7 support, all of which are cost-bearing features that likely justify premium pricing if utilized efficiently. | 中 | SI009, SI010, SI011, SI013 |
| CI027 | Proxyway’s Oxylabs review describes the company as one of the fastest and most reliable premium providers in its testing, which supports willingness-to-pay but does not reveal margin. | 中 | SI016, SI018, SI019 |
| CI028 | BestProxyCompare says proxy prices fell across the board in 2026 as competition intensified, indicating real market pressure on commodity layers. | 低 | SI023 |
| CI029 | Decodo’s public pricing page advertises residential proxies from $2/GB, materially below Oxylabs’ visible residential list-price bands. | 中 | SI022 |
| CI030 | These market-floor comparisons imply that Oxylabs’ financial case depends on premium performance and enterprise value-adds, not on being the cheapest provider. | 中 | SI016, SI022, SI023 |
| CI031 | The visible product and pricing mix suggests revenue quality should be relatively strong because multiple products monetize on recurring usage rather than one-off project fees. | 中 | SI009, SI010, SI011, SI013 |
| CI032 | No retained public source discloses revenue mix by product, realized discounting, gross margin, or net revenue retention. | 中 | SI001, SI002, SI007 |
| CI033 | No retained public source discloses cash on hand, monthly burn, or runway months after the July 2026 round. | 中 | SI001, SI002, SI008 |
| CI034 | No retained public source discloses debt facilities, project-finance obligations, or material covenant structures. | 中 | SI001, SI002 |
| CI035 | A decade of bootstrapping before taking outside capital implies low historical financing dependency. | 中 | SI002, SI003, SI004, SI005 |
| CI036 | The fresh $130 million round likely improved capital adequacy materially, but the exact runway impact cannot be modeled without cash and burn data. | 中 | SI001, SI002, SI008 |
| CI037 | Using $3.6 billion valuation and $350 million ARR implies an approximate valuation-to-ARR multiple of 10.3x. | 中 | SI001, SI002 |
| CI038 | Using 350,000 tech teams as the denominator implies roughly $1,000 of ARR per team, while using 15,000 clients implies roughly $23,300 of ARR per client. | 低 | SI001, SI007 |
| CI039 | Using 400-500 employees from Proxyway or 500+ experts from official disclosures implies a rough ARR-per-employee or expert range of about $700,000 to $875,000. | 低 | SI001, SI020 |
| CI040 | Because core operating metrics remain private, public valuation inputs are directionally useful but not sufficient for a full margin and cash model. | 中 | SI001, SI002, SI007, SI020 |
| CI041 | The absence of visible debt or project finance, together with the long bootstrapped history, supports a tentative read that Oxylabs is more capital-light than asset-heavy. | 中 | SI002, SI005, SI025 |
| CI042 | TNW’s warning that large-scale scraping and residential proxy sourcing still draw legal scrutiny means compliance cost and procurement friction remain part of the financial model. | 中 | SI003 |
| CI043 | Oxylabs AI Studio exposes separate public pricing, with plans starting around $12 per month and scaling toward custom enterprise packages, indicating an additional monetization layer above raw proxies. | 中 | SI026 |
| CI044 | Oxylabs’ dedicated Web Scraper API pricing page shows named monthly plans such as $49 micro, $99 starter, and $249 advanced, plus a free trial up to 2,000 results. | 中 | SI027 |
| CI045 | Oxylabs’ Fast Search API page positions the product for millions of queries per day and guaranteed uptime, supporting an AI-search workload monetization thesis even where exact list pricing is not shown. | 中 | SI028 |
| CI046 | Webshare’s own pricing page shows residential traffic as low as $1.40/GB at scale alongside free proxies, reinforcing how far the market floor sits below Oxylabs’ premium public pricing. | 中 | SI029 |
| CI047 | Proxyway’s 2026 Webshare review says the platform remains especially attractive to buyers prioritizing flexibility, affordability, and self-service, underscoring a real low-end pricing threat to premium vendors. | 中 | SI030 |
| CE001 | Oxylabs’ homepage frames the platform around AI training, RAG, and other workflow-level use cases rather than around proxy access alone. | 中 | SE001 |
| CE002 | The residential proxies page emphasizes advanced filtering, premium IP quality, and response-time optimization, including a public 0.41s response-time signal. | 中 | SE002 |
| CE003 | The datacenter proxies page advertises 99.9% success rate, 99.9% uptime, and HTTP/HTTPS/SOCKS5 support. | 中 | SE003 |
| CE004 | Oxylabs positions Web Scraper API as an all-in-one source of structured, ready-to-use data from multiple websites through one API. | 中 | SE004 |
| CE005 | Web Scraper API quick-start and product docs say the product automatically handles proxy rotation, request retries, rendering, and anti-bot systems. | 中 | SE011, SE013 |
| CE006 | The Web Scraper API quick start exposes source, URL or query, parse, render, and geo_location as core user controls. | 中 | SE011 |
| CE007 | The Web Scraper API quick start points users to realtime, proxy-endpoint, and asynchronous push-pull integration methods. | 中 | SE011 |
| CE008 | The Web Unblocker quick start says the product is an AI-powered proxy solution that automatically manages rotation, fingerprinting, and CAPTCHA handling for sophisticated bot defenses such as Akamai, Cloudflare, and DataDome. | 中 | SE012 |
| CE009 | Web Unblocker is operated through the unblock.oxylabs.io:60000 endpoint and headers for geo location, rendering, session persistence, and forced headers, with a 550 code when the system fails after retries. | 中 | SE012 |
| CE010 | Fast Search API is presented as a high-speed lightweight SERP scraper that can handle millions of queries per day for apps, agents, and LLM workflows. | 中 | SE006 |
| CE011 | AI-Scraper uses natural-language prompts, automatic schema generation, JSON or Markdown output, and optional JavaScript rendering instead of manual CSS/XPath selectors. | 中 | SE015 |
| CE012 | AI-Scraper docs say the tool can integrate through Python and JavaScript SDKs, an MCP server, or third-party integrations. | 中 | SE015 |
| CE013 | AI Studio has live pricing and plan packaging, which shows it is already commercialized rather than merely experimental. | 中 | SE008 |
| CE014 | The docs overview presents Oxylabs products, API targets, and integrations as one unified documentation surface. | 中 | SE010 |
| CE015 | Across official site and docs surfaces, the visible Oxylabs product family includes residential proxies, datacenter proxies, Web Scraper API, Web Unblocker, Fast Search API, and AI Studio apps. | 中 | SE001, SE004, SE005, SE006, SE007 |
| CE016 | Oxylabs’ product design trends upward from raw access toward progressively more managed and AI-shaped outputs. | 中 | SE004, SE006, SE007, SE015 |
| CE017 | Proxyway says proxy APIs combine multiple IP types, an intelligent proxy-management layer, and website-unblocking mechanisms. | 中 | SE017 |
| CE018 | Proxyway’s 2026 rankings keep Oxylabs Web Unblocker in the top tier for professional use but also compare it directly against other strong managed stacks. | 中 | SE018, SE019 |
| CE019 | Bright Data docs show a similarly comprehensive managed-unblocking feature set including CAPTCHA solving, fingerprinting, and custom header handling, so these capabilities are not unique to Oxylabs. | 中 | SE020, SE021 |
| CE020 | ScraperAPI docs expose premium pools, ultra-premium pools, and rendering parameters, showing that other vendors also abstract proxy complexity into higher-level controls. | 中 | SE022, SE023 |
| CE021 | Apify’s comparison says Bright Data offers 600+ ready-made scrapers and dataset-marketplace breadth, while Oxylabs leans more on a general API and playground model. | 中 | SE026 |
| CE022 | GitHub organization and quick-start-guide pages show that Oxylabs invests in public developer enablement alongside its docs and trials. | 中 | SE024, SE025 |
| CE023 | The Web Scraper API docs list a broad set of target-specific sources spanning Amazon, Google, YouTube, Walmart, eBay, Etsy, Best Buy, Instacart, Kroger, and more. | 中 | SE013 |
| CE024 | The Web Scraper API docs explicitly mention enterprise security standards including SOC 2 Type II compliance. | 中 | SE013 |
| CE025 | Web Unblocker controls include geo selection, HTML or PNG rendering, sticky sessions, custom-header forwarding, rate limits, and explicit error codes. | 中 | SE012, SE014 |
| CE026 | AI-Scraper exposes render_javascript and geo_location parameters alongside prompt, schema, and output-format controls. | 中 | SE015 |
| CE027 | Competitor-authored and third-party materials consistently frame Oxylabs managed products as enterprise-grade and premium rather than budget-oriented or beginner-first. | 中 | SE018, SE020, SE026 |
| CE028 | Decodo and Webshare demonstrate how much cheaper or simpler lower-end alternatives can look when buyers only need core proxy functionality. | 中 | SE027, SE028, SE029 |
| CE029 | Oxylabs’ critical dependencies include IP supply, target-site anti-bot behavior, rendering infrastructure, parser maintenance, docs quality, and support operations. | 中 | SE010, SE011, SE012, SE017 |
| CE030 | The impact-report launch of AI Studio and ScrapingBee, plus 2026 live pricing/docs surfaces and recently updated quick starts, show ongoing product iteration toward AI-native workflows. | 中 | SE008, SE011, SE012, SE015 |
| CE031 | Residential proxies, datacenter proxies, Web Scraper API, and Web Unblocker appear mature, while Fast Search API and especially AI Studio look newer but commercially active. | 中 | SE002, SE003, SE004, SE005, SE006, SE008, SE015 |
| CE032 | The visible customer workflows span any-public-website scraping, supported-domain structured extraction, AI-agent search, and natural-language extraction from single pages. | 中 | SE004, SE006, SE015 |
| CE033 | The strongest public differentiation case is platform coherence across layers rather than one evidently unique underlying mechanism. | 中 | SE010, SE017, SE021, SE026 |
| CE034 | Authentication, rate limits, error codes, compliance language, geo controls, and session controls show that trust and quality are implemented as explicit product controls. | 中 | SE012, SE013, SE014 |
| CE035 | Success-rate, uptime, retry, parser, and response-time language across Oxylabs product pages shows quality is marketed as a core product outcome. | 中 | SE002, SE003, SE004, SE005 |
| CE036 | Because the product depends on moving targets and external data-access conditions, docs and support are operational dependencies rather than mere documentation extras. | 中 | SE011, SE012, SE017, SE024 |
| CU001 | Oxylabs’ homepage says the company is trusted by 15,000+ clients globally. | 中 | SU001 |
| CU002 | Oxylabs’ July 2026 funding announcement says the platform is trusted by more than 350,000 tech teams worldwide. | 中 | SU006 |
| CU003 | Oxylabs’ case-studies hub says the company serves customers from mid-market to enterprise and across multiple industries. | 中 | SU002 |
| CU004 | Wiser’s case study describes a workflow gathering public e-commerce data multiple times per day across thousands of domains. | 中 | SU003 |
| CU005 | Wiser says Oxylabs’ datacenter proxies are instrumental to establishing a rock-solid foundation for its workflows. | 中 | SU003 |
| CU006 | Wiser’s reference says the partnership is expected to continue as Wiser expands its breadth of data collection and capabilities. | 中 | SU003 |
| CU007 | Epicup is an SEO software provider that uses Oxylabs’ SERP Scraper API, now part of Web Scraper API, for high-quality large-scale search data. | 中 | SU004 |
| CU008 | Epicup says fair pricing, low response time, and high average success rate were its three core vendor requirements. | 中 | SU004 |
| CU009 | The Environmental Protection Department deployment has been integrated into daily operations since July 2024 and scans roughly 400 ads per week. | 中 | SU005 |
| CU010 | The Environmental Protection Department case study says the deployment has supported 15 investigations and 4 confirmed violations while reducing manual review time to about 1 hour per week. | 中 | SU005 |
| CU011 | Named customer proof spans retail intelligence, SEO analytics, and public-sector enforcement rather than one narrow vertical. | 中 | SU003, SU004, SU005 |
| CU012 | The combination of 15,000+ global clients and cross-industry case studies suggests Oxylabs’ customer footprint is international rather than Lithuania-centric. | 中 | SU001, SU002 |
| CU013 | Proxyway describes Oxylabs as the best enterprise proxy provider and emphasizes scale, ethics, and support. | 中 | SU009 |
| CU014 | G2 shows Oxylabs at 4.5 out of 5 from 421 reviews and highlights reliable proxies, easy setup, and responsive support. | 中 | SU014 |
| CU015 | Capterra reviews emphasize easy setup, good support, stability, and spending transparency but also flag premium pricing as a barrier for some users. | 中 | SU015 |
| CU016 | Trustpilot review content is materially more mixed, including both strong praise for account managers and explicit complaints about refunds or poor support. | 中 | SU016 |
| CU017 | Gartner reviewers say Oxylabs is especially valuable for serious, high-volume enterprise scraping, while smaller operators may find better value elsewhere. | 中 | SU017 |
| CU018 | Thunderbit argues that Oxylabs’ customer conversation is more polarized in 2026, with enterprise buyers rating it highly and smaller operators focusing on price and friction. | 低 | SU019 |
| CU019 | No retained public source discloses Oxylabs NRR, GRR, churn, or renewal rates. | 中 | SU001, SU006, SU014, SU017 |
| CU020 | No retained public source discloses public contract length, term mix, or logo-retention cohorts. | 中 | SU001, SU006, SU017 |
| CU021 | Public satisfaction visibility is much stronger than public retention visibility. | 中 | SU014, SU015, SU016, SU017 |
| CU022 | The visible product stack implies a land-and-expand path from basic proxy access into Web Scraper API, Web Unblocker, and newer AI-oriented tools. | 中 | SU023, SU024, SU025 |
| CU023 | The Webshare acquisition preserved a self-serve customer path and therefore broadens channel breadth beyond Oxylabs’ premium enterprise brand alone. | 中 | SU008, SU022 |
| CU024 | Oxylabs’ AI and agentic positioning likely broadens the addressable customer set toward AI builders and developer workflows beyond legacy scraping teams. | 中 | SU006, SU007, SU023 |
| CU025 | Public sources are insufficient to quantify top-customer concentration or revenue contribution by segment. | 中 | SU001, SU006, SU017 |
| CU026 | Wiser and Epicup look like production customer proof, while the Environmental Protection Department is deployed and outcome-rich but not standard recurring-revenue proof. | 中 | SU003, SU004, SU005 |
| CU027 | KYC or verification friction is a visible customer risk for smaller or less enterprise-like buyers. | 中 | SU015, SU019 |
| CU028 | Smaller or price-sensitive users are more likely to question Oxylabs’ value proposition than enterprise-heavy users. | 中 | SU016, SU017, SU019, SU021, SU022 |
| CU029 | Oxylabs likely serves a mix of enterprise, mid-market, and long-tail self-serve users, but the public record does not show segment revenue weights. | 中 | SU002, SU008, SU014 |
| CU030 | Customer proof quality is moderate overall: the named proof is real and recent, but the set is small relative to the company’s claimed overall footprint. | 中 | SU002, SU003, SU004, SU005 |
| CU031 | Support is both a strength and a recurring complaint in public customer evidence, implying experience may vary materially by segment or issue type. | 中 | SU014, SU015, SU016 |
| CU032 | BestProxyCompare’s 2026 guidance that enterprises should shortlist Oxylabs while budget buyers may prefer other vendors supports a bifurcated customer-quality story. | 低 | SU020 |
| CU033 | Decodo and Webshare illustrate realistic low-cost substitutes for customers who do not need Oxylabs’ premium enterprise positioning. | 中 | SU021, SU022 |
| CU034 | The public adoption trajectory includes some credible deployment metrics and review volume, but not active paid deployments by segment or product. | 中 | SU003, SU005, SU014, SU015, SU017 |
| CU035 | Expansion likely comes from deeper workflow embedding and cross-sell into broader product layers, while concentration risk likely increases as accounts become more enterprise-heavy. | 中 | SU023, SU024, SU025, SU020 |
| CU036 | The customer verdict is favorable on breadth and proof of use, but still incomplete on durability because retention and concentration metrics remain private. | 中 | SU003, SU014, SU017, SU019 |
| CU037 | G2's Oxylabs product page independently features a Wiser customer-proof asset, reinforcing that Wiser is used publicly as a reference account for Oxylabs. | 低 | SU014 |
| CR001 | Oxylabs operates a public trust center that positions security, privacy, and ethical compliance as core parts of its commercial posture. | 中 | SR001, SR002 |
| CR002 | Oxylabs says Scraper API and Web Unblocker achieved SOC 2 Type 2 certification. | 高 | SR001, SR011 |
| CR003 | Oxylabs publicly surfaces ISO 27001:2022 certification through its trust materials. | 中 | SR001, SR011 |
| CR004 | Oxylabs’ trust materials explicitly frame ethical sourcing and fair compensation of residential-proxy participants as a trust question, implying continued dependence on source-quality integrity. | 中 | SR001 |
| CR005 | Oxylabs’ KYC policy says every customer must answer a KYC questionnaire before gaining access to solutions. | 中 | SR003 |
| CR006 | Oxylabs reserves the right to refuse service and says illegal or unethical use is subject to refusal or immediate contract termination. | 中 | SR003 |
| CR007 | Oxylabs’ general conditions limit service use to lawful and legitimate purposes and prohibit competitive monitoring of service availability or performance without written consent. | 中 | SR004 |
| CR008 | Oxylabs promises only reasonable efforts to make services available 24/7 and explicitly carves out downtime tied to planned maintenance, government action, ISP failure, or denial-of-service attacks. | 中 | SR004 |
| CR009 | Oxylabs’ acceptable use policy acknowledges that automated data gathering is not automatically illegal but can become unlawful depending on method and use case. | 中 | SR005 |
| CR010 | Oxylabs may suspend or terminate customer service for AUP violations, including on a first offense. | 中 | SR005 |
| CR011 | Oxylabs’ DPA allows subprocessors and says customers may request the current subprocessor list in writing. | 中 | SR006 |
| CR012 | Oxylabs’ DPA requires prompt notification of personal-data breaches and return or deletion of personal data at the end of processing. | 中 | SR006 |
| CR013 | GDPR is the governing EU regulation for personal-data processing and therefore provides the baseline legal frame for any scraping workflow that processes personal data. | 中 | SR009 |
| CR014 | The EDPB’s July 2026 web-scraping guidance says GDPR applies to web scraping when it includes personal data processing and emphasizes data minimisation and legitimate-interest analysis. | 高 | SR009, SR010 |
| CR015 | EU privacy compliance risk for Oxylabs-adjacent workflows is rising because public availability does not eliminate GDPR obligations when scraped content includes personal data. | 高 | SR009, SR010 |
| CR016 | Article 50 transparency obligations under the EU AI Act apply from 2 August 2026 and can matter to AI-oriented extraction or customer-facing AI features. | 中 | SR031, SR008 |
| CR017 | Commentary on Meta v. Bright Data indicates that a favorable public-data scraping outcome in one case does not automatically make scraping universally lawful. | 中 | SR012, SR013 |
| CR018 | LinkedIn’s user agreement broadly prohibits automated scraping, copying, and circumvention, reinforcing endpoint-specific contract risk for some customer workflows. | 中 | SR032 |
| CR019 | The legal perimeter for public-web data collection remains target-specific because permissive public-data case law and restrictive platform terms can coexist. | 中 | SR009, SR010, SR013, SR032 |
| CR020 | Cloudflare’s pay-per-crawl feature lets site owners monetize AI crawler access and can return HTTP 402 Payment Required responses with pricing. | 中 | SR007 |
| CR021 | TechCrunch reports that from 15 September 2026 Cloudflare will block mixed-use crawlers by default on ad-hosting pages for new sites and free customers. | 高 | SR007, SR022 |
| CR022 | Publisher monetization and default blocking policies can increase Oxylabs customer cost-to-serve and reduce technical success rates on some target classes. | 中 | SR007, SR022, SR029, SR030 |
| CR023 | Independent review surfaces consistently position Oxylabs as strongest for enterprise or high-volume users rather than for the smallest budget-conscious operators. | 中 | SR014, SR017, SR018, SR019 |
| CR024 | Trustpilot and Thunderbit indicate more mixed sentiment among smaller or less enterprise-like users, especially around refunds, support, and pricing friction. | 中 | SR015, SR016 |
| CR025 | Webshare’s free tier and Decodo’s pay-as-you-go offer illustrate active lower-cost substitutes that can pressure Oxylabs pricing outside the highest-value enterprise cohorts. | 中 | SR020, SR021 |
| CR026 | Oxylabs’ July 2026 funding narrative links the business to 350,000+ tech teams and AI-era web-data demand, which increases exposure to AI-related regulatory and platform shifts. | 中 | SR023 |
| CR027 | Named customer references from Wiser, Epicup, and the Environmental Protection Department show production-like or embedded workflow use rather than only pilot usage. | 中 | SR024, SR025, SR026 |
| CR028 | Public sources still do not disclose Oxylabs NRR, gross margin, cash burn, or top-customer concentration. | 中 | SR023, SR028, SR014 |
| CR029 | Oxylabs’ published trust, KYC, contractual, and privacy materials constitute meaningful mitigations relative to less formal proxy-market peers. | 中 | SR001, SR003, SR004, SR005, SR006 |
| CR030 | No public breach or major operational incident surfaced in the fetched source set, but that absence does not remove residual reliability or security risk. | 中 | SR001, SR002, SR006 |
| CR031 | Oxylabs’ Web Unblocker and Web Scraper API product promises imply that customers depend on Oxylabs to preserve success rates and structured-data delivery as target conditions change. | 中 | SR029, SR030 |
| CR032 | Taking first-time institutional capital after a long bootstrapped run likely raises governance, reporting, and execution expectations even if it reduces funding risk. | 中 | SR023 |
| CR033 | Webshare broadens Oxylabs’ acquisition surface toward lower-friction self-serve users, but it also likely broadens abuse, onboarding, and support-complexity risk. | 中 | SR020, SR027 |
| CR034 | Gartner specifically suggests that smaller operators may find better value elsewhere, reinforcing the risk of a segmented customer experience. | 中 | SR014 |
| CR035 | Trustpilot review content shows that infrastructure-provider framing and refund rules can create expectation mismatch for some users. | 中 | SR015 |
| CR036 | Because named customers use Oxylabs inside recurring operational workflows, any major reliability or compliance problem would likely transmit directly into customer outcomes rather than remain abstract. | 中 | SR024, SR025, SR026 |
| CR037 | Oxylabs’ own service agreement anticipates downtime from external causes such as government action, ISP failure, and denial-of-service attacks, underscoring that some availability risk is outside direct company control. | 中 | SR004 |
| CR038 | The absence of a publicly listed subprocessor roster, incident history, or customer-level SLA outcomes leaves a material diligence gap even though some of those artifacts may be available privately. | 中 | SR001, SR006 |
| CR039 | The DPA’s audit and verification rights help enterprise procurement, but they also imply customers will expect operational evidence beyond marketing claims. | 中 | SR006 |
| CR040 | Oxylabs’ top residual risks are regulatory tightening, target-site and publisher blocking, supply / trust dependencies, and opaque economics rather than a lack of market demand. | 中 | SR010, SR022, SR023, SR028, SR029, SR030 |
| CR041 | No material public enforcement action or lawsuit against Oxylabs itself surfaced in the fetched source set. | 低 | SR001, SR002, SR012, SR013 |
| CR042 | The Meta / LinkedIn contrast shows that customer legal risk cannot be evaluated at the category level alone; it must be tested against the specific endpoints a workflow targets. | 中 | SR013, SR032 |
| CV001 | Oxylabs raised $130 million in July 2026 at an approximately $3.6 billion valuation in its first outside round. | 高 | SV001, SV002, SV003 |
| CV002 | Independent July 2026 coverage says Oxylabs was running at roughly $350 million ARR and more than 350,000 customers at the time of the round. | 高 | SV003, SV004, SV005 |
| CV003 | Oxylabs’ own funding materials emphasize that the business spent a decade growing profitably without external capital before the Warburg investment. | 中 | SV001 |
| CV004 | Warburg and Oxylabs position the company as infrastructure for AI and agentic applications rather than only as a proxy vendor. | 中 | SV001, SV002 |
| CV005 | Using the public round valuation of $3.6 billion and roughly $350 million ARR implies an approximately 10.3x ARR multiple. | 高 | SV002, SV003, SV005 |
| CV006 | Similarweb traded at about 2.28x trailing sales in July 2026. | 中 | SV006 |
| CV007 | ZoomInfo traded at about 0.73x trailing sales in July 2026. | 中 | SV007 |
| CV008 | Cloudflare traded at about 45.7x trailing sales in July 2026. | 中 | SV008 |
| CV009 | C3.ai traded at about 4.60x trailing sales in July 2026. | 中 | SV009 |
| CV010 | Palantir traded at about 61.7x trailing sales in July 2026. | 中 | SV010 |
| CV011 | ZoomInfo’s official Q1 2026 results showed $310.2 million of revenue with only 1.5% year-over-year growth but 35% adjusted operating margin. | 中 | SV011 |
| CV012 | Similarweb’s official Q1 2026 results showed $73.9 million of revenue, 10% growth, positive non-GAAP operating profit, and multi-year subscriptions equal to 64% of ARR. | 中 | SV012 |
| CV013 | Cloudflare’s official Q1 2026 results showed $639.8 million of revenue, 34% growth, and 71.2% GAAP gross margin. | 中 | SV013 |
| CV014 | C3.ai’s fiscal 2026 SEC results showed $250.3 million of revenue, $575.4 million of cash and investments, and ongoing restructuring to reduce cost burn. | 中 | SV014 |
| CV015 | Palantir’s Q1 2026 10-Q disclosed gross margins around 87% and 80%, supporting why its public premium is much higher than Oxylabs’ implied multiple. | 中 | SV015 |
| CV016 | Similarweb, ZoomInfo, C3.ai, and Palantir all provide filing or filing-adjacent public disclosures that let investors inspect revenue and risk factors more directly than Oxylabs does today. | 中 | SV014, SV015, SV016, SV017, SV018, SV019 |
| CV017 | The public-comp band for Oxylabs is extremely wide, ranging from sub-1x sales for challenged data vendors to 40x-60x for premium AI or internet-infrastructure leaders. | 中 | SV007, SV008, SV009, SV010 |
| CV018 | Independent reviews position Oxylabs as a premium provider especially suitable for enterprise or high-volume use cases. | 中 | SV020, SV023, SV024, SV025 |
| CV019 | Adverse review sources suggest smaller or more price-sensitive users may experience Oxylabs less favorably than enterprise-heavy cohorts. | 中 | SV021, SV022 |
| CV020 | Wiser, Epicup, and the Environmental Protection Department show that Oxylabs is embedded in recurring operational workflows, not only in lightweight trials. | 中 | SV027, SV028, SV029 |
| CV021 | Oxylabs uses multiple public customer-scale definitions, including 15,000+ clients on its homepage and 350,000+ customers or tech teams in round-related materials. | 中 | SV001, SV003, SV026 |
| CV022 | Customer-count and ARR-definition ambiguity weakens confidence in precise underwriting at the current valuation. | 中 | SV001, SV003, SV005, SV026 |
| CV023 | Public sources do not disclose Oxylabs NRR, gross margin, cash burn, or top-customer concentration. | 中 | SV001, SV002, SV003, SV026 |
| CV024 | Because the key economics are missing, current public evidence does not support an outright buy recommendation at the July 2026 price. | 中 | SV001, SV003, SV005, SV023 |
| CV025 | A bear-case valuation band of roughly $1.8B-$2.5B can be justified by applying about 5x-7x ARR to the public $350M ARR anchor. | 中 | SV003, SV005, SV006, SV007, SV009 |
| CV026 | A base-case valuation band of roughly $3.0B-$4.0B can be justified by applying about 8.5x-11.5x ARR to the public $350M ARR anchor. | 中 | SV003, SV005, SV009, SV018, SV023 |
| CV027 | A bull-case valuation band of roughly $4.2B-$5.6B can be justified if Oxylabs privately proves strong retention and software-like economics. | 中 | SV003, SV005, SV008, SV010 |
| CV028 | The July 2026 $3.6B round sits inside the base-case band implied by current public evidence. | 中 | SV002, SV003, SV005 |
| CV029 | The best current recommendation is track, with medium confidence, medium-high risk, and a fair valuation stance. | 中 | SV005, SV018, SV023 |
| CV030 | If private diligence confirms strong margins, healthy NRR, and low concentration, upside exists even without Cloudflare- or Palantir-like multiples. | 中 | SV005, SV018, SV020, SV023 |
| CV031 | If regulatory or platform friction materially raises cost to serve, the appropriate multiple could compress sharply toward the lower public-comp range. | 中 | SV021, SV022, SV030 |
| CV032 | Cloudflare’s September 2026 mixed-use crawler blocking policy reinforces a concrete downside trigger for web-data infrastructure valuations. | 中 | SV030 |
| CV033 | A decade of bootstrapped, profitably grown operating history suggests Oxylabs may deserve a premium to slower public data vendors that lack similar strategic momentum. | 中 | SV001, SV003, SV005 |
| CV034 | The Warburg round may embed some scarcity premium because it is the first outside capital into a rare, scaled, private European web-data-infrastructure asset. | 中 | SV002, SV003, SV005 |
| CV035 | No public comparable is a clean like-for-like match for Oxylabs, so public multiples should be used as boundary markers rather than as a single precise answer. | 中 | SV006, SV007, SV008, SV009, SV010 |
| CV036 | Similarweb is the most useful adjacent lower-band comp because it also monetizes digital data and analytics with AI-tailwind language, but it is still not a proxy-infrastructure twin. | 中 | SV006, SV012, SV016 |
| CV037 | ZoomInfo, C3.ai, Cloudflare, and Palantir together frame how public markets reward very different combinations of growth, margin, and narrative quality. | 中 | SV007, SV008, SV009, SV010, SV011, SV013, SV014, SV015 |
| CV038 | Before a firm buy call, investors need an audited ARR bridge, retention cohorts, gross-margin decomposition, and concentration analysis. | 中 | SV001, SV002, SV003, SV005 |
| CV039 | The thesis breaks if ARR quality is weaker than implied or if future round terms create unfavorable preference or dilution overhang. | 中 | SV001, SV002, SV003 |
| CV040 | The thesis also breaks if growth or customer durability disappoint enough to push the business toward the Similarweb/C3.ai end of the multiple range rather than the premium infrastructure end. | 中 | SV006, SV009, SV020, SV021, SV022, SV030 |
| CV041 | The overall evidence supports business quality and relevance, but not enough price dislocation to justify an aggressive buy recommendation at the current public anchor. | 中 | SV005, SV018, SV023, SV030 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | Oxylabs | Oxylabs - High Quality Proxy Service to Gather Data at Scale | Trusted by 15,000+ clients globally. |
| SO002 | Oxylabs | Oxylabs Reaches $3.6B Value After Agreeing to Its First Outside Investment | Oxylabs is recording $350 million in ARR as its data platform, trusted by over 350,000 tech teams worldwide. |
| SO003 | Warburg Pincus | Oxylabs Announces $130 Million Investment from Warburg Pincus | Values Oxylabs at approximately $3.6 billion. |
| SO004 | SiliconANGLE | Web data scraping infrastructure startup Oxylabs reels in $130M in its first funding round | |
| SO005 | The Next Web | Lithuania’s Oxylabs raises $130M from Warburg Pincus at a $3.6bn valuation | The web-data business carries baggage. Large-scale scraping sits in a legal grey zone. |
| SO006 | Tech Funding News | Oxylabs becomes Lithuania’s second most valuable private company at $3.6B | |
| SO007 | EU-Startups | New Lithuanian unicorn: Oxylabs ends bootstrapped streak after securing €113.6 million at €3.1 billion valuation | |
| SO008 | Business Insider / Markets Insider | Oxylabs Announces $130 Million Investment from Warburg Pincus | |
| SO009 | International Business Times | Oxylabs Reaches $3.6B Valuation With Warburg Pincus Investment | |
| SO010 | Tesonet | Oxylabs becomes the second bootstrapped Unicorn built in the Tesonet ecosystem | With over 350,000 customers and $350 million in ARR, Oxylabs has established itself as a critical enabler of large-scale data acquisition. |
| SO011 | Oxylabs | Oxylabs Group Evolves Its Leadership Structure to Accelerate the Next Decade | Julius Černiauskas, our founder and CEO, is transitioning to Chairman of the Board. |
| SO012 | Oxylabs | Oxylabs 2025 Impact Report Highlights a Decade of Public Web Data Leadership | The company launched Oxylabs AI Studio, acquired France-based web scraping API ScrapingBee, expanded its patent portfolio to 150, and introduced the Oxylabs Trust Center. |
| SO013 | Oxylabs | Risk and Legal Compliance | At Oxylabs, compliance is embedded in how we operate. |
| SO014 | Oxylabs | Oxylabs Trust Center | |
| SO015 | Oxylabs | Code of Ethics | |
| SO016 | Oxylabs | Buy Fast Residential IP Proxies From Best Provider - Free Trial | 175M+ best reputation IPs. |
| SO017 | Oxylabs | Buy Best 5G/4G/3G/LTE Rotating Mobile IP Proxies - Free Trial | 20M+ mobile IPs worldwide. |
| SO018 | Oxylabs | Buy Datacenter Proxies from $1.20/IP - Free Trial | Get the right Datacenter type for your data collection, all delivering a 99.9% success rate, 99.9% uptime, and consistently fast response times. |
| SO019 | Oxylabs | Web Scraper API: Start Web Data Collection - Free Trial | Integrate Web Scraper API with AI platforms. |
| SO020 | Oxylabs | Oxylabs Available Scraping Sources | |
| SO021 | Webshare | Joining forces with Oxylabs | Today, we are announcing that Oxylabs has completed its acquisition of Webshare Software. |
| SO022 | Oxylabs | Case Studies: Success Stories | |
| SO023 | G2 | The G2 on Oxylabs | Reviews 4.5; 421 reviews. |
| SO024 | Capterra | Oxylabs Reviews 2025. Verified Reviews, Pros & Cons - Capterra | The only downside is their premium pricing, which might be a barrier for smaller businesses or individual users. |
| SO025 | Trustpilot | Oxylabs is rated "Great" with 3.9 / 5 on Trustpilot | Be cautious with this company. Support is abysmal, and they do not refund with their apparent 1-click checkout process. |
| SO026 | Oxylabs | Oxylabs AI Predictions 2026: Five-Year Track of Early Calls and What the Future Holds | |
| SO027 | Oxylabs | DEVWorld 2026: Oxylabs on Web Scraping Infrastructure for AI | |
| SO028 | Oxylabs | Oxylabs Heads to the AI Engineer World’s Fair 2026 | |
| SO029 | Craft | Oxylabs Company Profile - Office Locations, Competitors, Revenue, Financials, Employees, Key People, Subsidiaries | Craft.co | Type Private Status Active Founded 2012 HQ Vilnius, LT. |
| SM001 | Oxylabs | Oxylabs AI Predictions 2026: Five-Year Track of Early Calls and What the Future Holds | |
| SM002 | Oxylabs | Risk and Legal Compliance | At Oxylabs, compliance is embedded in how we operate. |
| SM003 | Mordor Intelligence | Web Scraping Market Size, Growth Report, Share & Trends 2026 - 2031 | The Web Scraping Market worth USD 1.56 billion in 2026 is growing at a CAGR of 17.39% to reach USD 3.49 billion by 2031. |
| SM004 | The Business Research Company | Web Scraping Market Size and Growth Analysis Report 2026 | The web scraping market size will grow from $0.99 billion in 2025 to $1.17 billion in 2026. |
| SM005 | Cybernews | Proxyway Releases 2026 Proxy Server Market Report | Major players have been growing by 50% or more year-over-year. |
| SM006 | Apify Blog | State of web scraping report 2026 | 65.8% of respondents reported increased proxy usage, and 58.3% said their proxy spending increased year over year. |
| SM007 | ScrapingDog | Web Scraping Statistics & Trends You Need to Know in 2026 | 2026: web-scraping market racing toward multi-billion ($2.2–3.5B); alt-data at $4.9B, +28% YoY. |
| SM008 | WorldMetrics | Web Scraping Industry Statistics | Fact-Checked 2026 | 75% of enterprises use web scraping for competitive intelligence. |
| SM009 | OpenWebSearch.eu | Welcome - OpenWebSearch.eu – Promoting Europe‘s Independence in Web Search | 9.14 Billion URLs crawled. |
| SM010 | OpenWebSearch.eu | Open Web Index - OpenWebSearch.eu – Promoting Europe‘s Independence in Web Search | The index stores well structured open web data, making it available for search applications and LLMs. |
| SM011 | OpenWebSearch.eu | 42 months in the making: OpenWebSearch.EU project end results look promising! | |
| SM012 | OpenWebSearch.eu | Register for #ossym 2026 now! | |
| SM013 | Cloudflare Blog | Introducing pay per crawl: Enabling content owners to charge AI crawlers for access | Publishers will then have three distinct options for a crawler: Allow, Charge, Block. |
| SM014 | Cloudflare Docs | Pay Per Crawl · Cloudflare AI Crawl Control docs | |
| SM015 | Cloudflare Docs | What is Pay Per Crawl? · Cloudflare AI Crawl Control docs | |
| SM016 | EUR-Lex | Regulation - 2016/679 - EN - gdpr | |
| SM017 | EUR-Lex | Regulation - 2022/2065 - EN - Digital Services Act | |
| SM018 | EUR-Lex | Regulation - EU - 2024/1689 - EN | |
| SM019 | DataDome | Web Scraping Protection: How to Prevent Web Scraping - DataDome | Web scraping costs businesses billions annually. |
| SM020 | DataDome | The Agentic Threats & Industry Trends Defining 2026 (So Far) | 80% of AI agents don’t properly identify themselves when visiting websites. |
| SM021 | Proxyway | The Best Residential Proxies of 2026: Tested & Ranked - Proxyway | Oxylabs 99.93% success rate; 0.60 s response time. |
| SM022 | Proxyway | Oxylabs Proxies: In-Depth Review & Performance Tests | |
| SM023 | AIMultiple | 10 Best Proxy Providers for 2026: Tested & Ranked | As anti-bot systems become more sophisticated, the focus has shifted from raw IP volume to IP reputation. |
| SM024 | AIMultiple | Oxylabs vs Bright Data: Pricing & Benchmark Comparison | Oxylabs maintained consistently high success rates throughout the testing period, generally ranging between 90% and 98% for mobile proxies. |
| SM025 | AIMultiple | How Much Does a Proxy Cost? 2026 Proxy Pricing Comparison | On average, residential proxy prices range from $3 to $15 per GB. |
| SM026 | OpenWebIndex.eu | Open Web Index - Your daily dose of web data | |
| SM027 | European Commission | AI Act | High-risk AI systems are subject to strict obligations before they can be put on the market. |
| SP001 | Oxylabs | Oxylabs home page | Trusted by 15,000+ clients globally. |
| SP002 | Oxylabs | Residential proxies | |
| SP003 | Oxylabs | Web Unblocker | Ensure high success rates with Oxylabs’ dynamic scraping strategy discovery system. |
| SP004 | Bright Data | Bright Data vs Oxylabs Comparison 2026 | Bright Data: 400M+ IPs ... Oxylabs: 175M+ IPs. |
| SP005 | Bright Data | The 12 Best Proxy Providers of 2026 | Bright Data is the most complete proxy provider on the market. |
| SP006 | Bright Data | Best Web Unblockers 2026: Tested Comparison & Reviews | Bright Data Web Unlocker – 97.9% success rate ... Oxylabs Web Unblocker ... starts at $9.4/GB. |
| SP007 | Bright Data | Residential proxies | Over 400M+ monthly ethical residential IPs from 195 countries. |
| SP008 | Decodo | Decodo home page | AI-Ready Proxy & Scraping Solutions. |
| SP009 | Decodo | Residential proxies | Residential Proxies: 115M+ Ethically-Sourced IPs in 195+ Locations. |
| SP010 | Decodo | Pricing | Best value for money ... Pay As You Go ... 3-day free trial. |
| SP011 | SOAX | SOAX home page | SOAX handles IP rotation and routing in one connected system. |
| SP012 | SOAX | Pricing | No sales theatre. No pricing games. Just real infrastructure economics. |
| SP013 | Apify Blog | Oxylabs vs. Bright Data for web scraping | Bright Data offers a collection of 600+ ready-made scrapers ... Oxylabs offers a general scraper API. |
| SP014 | Apify | Apify home page | 53,640 tools for your AI. |
| SP015 | Webshare | Webshare home page | Get 10 Proxies for free ... 99.7% Success rate. |
| SP016 | ScraperAPI | Plans & Billing | ScraperAPI offers a free plan of 1,000 free API credits per month. |
| SP017 | IPRoyal | Residential proxies | Pay As You Go. |
| SP018 | Proxyway | The Best Residential Proxies of 2026 | Decodo ... Best value provider. Oxylabs ... Best premium residential proxy provider. |
| SP019 | AIMultiple | 10 Best Proxy Providers for 2026: Tested & Ranked | The focus has shifted from raw IP volume to IP reputation. |
| SP020 | Rayobyte | Proxy pricing | For residential proxies, getting a free trial is as simple as creating an account. |
| SP021 | ScrapingBee | Best rotating and residential proxies for web scraping in 2026 | If you need fine-grained control over raw proxies, other providers might be better, but ScrapingBee is recommended for teams ... wanting a managed experience. |
| SP022 | Webshare Help Center | Do you offer proxies for free? | We're excited to offer you 10 datacenter proxies free of charge with Webshare! |
| SP023 | PrivateProxyGuide | Best Bright Data Alternatives in 2026: Top Proxy Tools | Bright Data can feel expensive, complex, or overpowered for teams that only need one specific thing. |
| SP024 | BestProxyCompare | 18 Best Proxy Providers in 2026, Ranked & Compared | Cheapest Proxies is our #1 pick for 2026. |
| SP025 | Distill Intelligence | Bright Data competitors: complete list | Leading competitors to Bright Data in 2026 include Apify, Oxylabs, and SOAX. |
| SI001 | Oxylabs | Oxylabs Reaches $3.6B Value After Agreeing to Its First Outside Investment | Oxylabs is recording $350 million in ARR ... trusted by over 350,000 tech teams worldwide. |
| SI002 | Warburg Pincus | Oxylabs Announces $130 Million Investment from Warburg Pincus | Marks Oxylabs’ first outside investment since its founding in 2015. |
| SI003 | The Next Web | Lithuania’s Oxylabs raises $130M from Warburg Pincus at a $3.6bn valuation | The web-data business carries baggage. Large-scale scraping sits in a legal grey zone. |
| SI004 | EU-Startups | New Lithuanian unicorn: Oxylabs ends bootstrapped streak after securing €113.6 million at €3.1 billion valuation | The company currently serves more than 350,000 customers and generates €305.8 million ($350 million) in annual recurring revenue (ARR). |
| SI005 | Tesonet | Oxylabs Secures First Outside Investment from Warburg Pincus, Reaching $3.6 Billion Valuation and Becoming 2nd Bootstrapped Unicorn in Tesonet Portfolio | The investment marks Oxylabs’ first outside funding since its founding in 2015. |
| SI006 | Oxylabs | Oxylabs 2025 Impact Report Highlights a Decade of Public Web Data Leadership | The company launched Oxylabs AI Studio, acquired France-based web scraping API ScrapingBee, expanded its patent portfolio to 150, and introduced the Oxylabs Trust Center. |
| SI007 | Oxylabs | Oxylabs home page | Trusted by 15,000+ clients globally. |
| SI008 | Business Wire / Markets Insider syndication | Oxylabs Announces $130 Million Investment from Warburg Pincus | The investment is being made by the Warburg Pincus Capital Solutions Founders Fund (“WPCS FF”). |
| SI009 | Oxylabs | Buy Fast Residential IP Proxies From Best Provider - Free Trial | Starter ... 5GB ... $6 /GB ... Corporate ... $2.50 ... billed monthly. |
| SI010 | Oxylabs | Buy Datacenter Proxies from $1.20/IP - Free Trial | Buy Datacenter Proxies from $1.20/IP - Free Trial. |
| SI011 | Oxylabs | Web Scraper API: Start Web Data Collection - Free Trial | Amazon: $0.25/1K results ... Google: $0.50/1K results ... Other: $0.70/1K results. |
| SI012 | Oxylabs | Oxylabs Available Scraping Sources | Explore ready-to-use Web Scraper API sources for popular websites across e-commerce, travel, real estate, AI platforms, and more. |
| SI013 | Oxylabs | Web Unblocker | Micro ... $9.40/GB ... Starter ... $8.60/GB ... Advanced ... $7.50/GB. |
| SI014 | Webshare | Joining forces with Oxylabs | No Price Change ... we continue our commitment to offering the most competitive proxy products at the lowest prices. |
| SI015 | Proxyway | Oxylabs Receives $130 Million Investment from Warburg Pincus | The investment comes from the Warburg Pincus Capital Solutions Founders Fund. |
| SI016 | Proxyway | Oxylabs Proxies: In-Depth Review & Performance Tests | Residential proxies are Oxylabs’ main proxy product ... sourced via Honeygain, direct partnerships with ISPs, and app developers. |
| SI017 | Proxyway | Search APIs in 2026: Overview & Benchmarks | Established providers like Oxylabs, Bright Data, and SerpApi have since introduced fast versions of their own. |
| SI018 | Proxyway | The Best Residential Proxies of 2026 | Oxylabs ... Best premium residential proxy provider. |
| SI019 | Proxyway | 11 Best ISP (Static Residential) Proxies of 2026 | Here, we’ll highlight the infrastructure’s success rate and response time as the main metrics. |
| SI020 | Proxyway | Proxy Market Research 2026 | Perhaps the market leader, competing with Bright Data in the enterprise segment. Oxylabs sources IPs through partner apps Honeygain, JumpTask, and an SDK. |
| SI021 | ScraperAPI Docs | Plans & Billing | ScraperAPI offers a free plan of 1,000 free API credits per month. |
| SI022 | Decodo | Pricing | Residential proxies from just $2/GB, no coupon needed. |
| SI023 | BestProxyCompare | 18 Best Proxy Providers in 2026, Ranked & Compared | Prices fell across the board as competition intensified. |
| SI024 | SiliconANGLE | Web data scraping infrastructure startup Oxylabs reels in $130M in its first funding round | |
| SI025 | International Business Times | Oxylabs Reaches $3.6B Valuation With Warburg Pincus Investment | The company says it now records $350 million in annual recurring revenue and is trusted by more than 350,000 technology teams worldwide. |
| SI026 | Oxylabs AI Studio | AI Studio Pricing – Affordable AI Data Extraction Apps | Starter: $12 per month ... Lite: $62 ... Standard: $250 ... Custom/Scalable: starts from $1,200 per month. |
| SI027 | Oxylabs | Web Scraper API pricing - Free Trial | Micro Plan ... $49 monthly ... Starter ... $99 monthly ... Advanced ... $249 monthly. |
| SI028 | Oxylabs | Buy Fast Search API From the Best Provider | Fast Search API is a high-speed and lightweight SERP scraper, easily handling millions of queries per day with guaranteed uptime. |
| SI029 | Webshare | Proxy Pricing - Check Latest Proxy Price | Rotating Residential Proxies: pricing as low as $1.40/GB when buying in large volumes. |
| SI030 | Proxyway | Webshare Proxies: In-Depth Review & Performance Tests | Webshare continues to stand out ... especially for users prioritizing flexibility, affordability, and self-service options. |
| SE001 | Oxylabs | Oxylabs - High Quality Proxy Service to Gather Data at Scale | Built for every use case - powered by one platform. |
| SE002 | Oxylabs | Buy Fast Residential IP Proxies From Best Provider - Free Trial | Get the best IPs, picked by our advanced filtering algorithm to ensure top performance. |
| SE003 | Oxylabs | Buy Datacenter Proxies from $1.20/IP - Free Trial | Get the right Datacenter type for your data collection, all delivering a 99.9% success rate, 99.9% uptime. |
| SE004 | Oxylabs | Web Scraper API: Start Web Data Collection - Free Trial | Structured, ready-to-use data from multiple websites through one API. |
| SE005 | Oxylabs | Web Unblocker | Ensure high success rates with Oxylabs’ dynamic scraping strategy discovery system. |
| SE006 | Oxylabs | Buy Fast Search API From the Best Provider | Fast Search API is a high-speed and lightweight SERP scraper, easily handling millions of queries per day. |
| SE007 | Oxylabs AI Studio | AI Studio – AI Web Scraping & Data Extraction Apps | Extract data from any webpage with AI. Describe your needs in plain English. |
| SE008 | Oxylabs AI Studio | AI Studio Pricing – Affordable AI Data Extraction Apps | Start small or scale to millions – pick the plan that fits your data needs. |
| SE009 | Oxylabs AI Studio | API Documentation | Explore the API documentation for all Oxylabs AI Studio apps. |
| SE010 | Oxylabs Docs | Overview | Oxylabs Documentation | Getting Started ... Products ... API Targets ... Integrations. |
| SE011 | Oxylabs Docs | Quick Start: Web Scraper API | Web Scraper API is an all-in-one solution that handles proxy rotation, JavaScript rendering, and anti-bot systems automatically. |
| SE012 | Oxylabs Docs | Quick Start: Web Unblocker | Web Unblocker is an AI-powered proxy solution designed to handle the most sophisticated bot traffic management systems. |
| SE013 | Oxylabs Docs | Web Scraper API | Products | The tool is built to meet enterprise security standards, including SOC 2 Type II compliance. |
| SE014 | Oxylabs Docs | Web Unblocker | Products | |
| SE015 | Oxylabs Docs | AI-Scraper | Products | This AI scraper removes the need for CSS/XPath selectors or custom parsers. |
| SE016 | Oxylabs Blog | How to Use Web Scraper API: Quick Start Guide | The Web Scraper API is a data collection platform that handles crawling and parsing, delivering data to cloud storage. |
| SE017 | Proxyway | In-Depth Look into Popular Proxy APIs (Web Unblockers) | The API combines multiple IP types, an intelligent proxy management layer, and website unblocking mechanisms. |
| SE018 | Proxyway | The Best Web Unblockers of 2026: Tested and Ranked | Oxylabs Web Unblocker ... top pick for the pros. |
| SE019 | Proxyway | Comparing Popular Web Scraping & Proxy APIs | |
| SE020 | Bright Data | Best Web Unblockers 2026: Tested Comparison & Reviews | Oxylabs Web Unblocker ... Starts at $9.4/GB. |
| SE021 | Bright Data Docs | Web Unlocker API features | By default ... Web Unlocker API also solves CAPTCHAs ... customizing headers, fingerprinting, and more. |
| SE022 | ScraperAPI Docs | Premium Residential/Mobile Proxy Pools | Requests using the premium=true parameter are charged at 10x the normal API credit rate. |
| SE023 | ScraperAPI Docs | Supported Parameters | |
| SE024 | GitHub | oxylabs organization on GitHub | Pinned ... Web scraping, data parsing and automation tutorials. |
| SE025 | GitHub | oxylabs/quick-start-guide | Search Engines ... E-Commerce Websites ... Other Websites. |
| SE026 | Apify Blog | Oxylabs vs. Bright Data for web scraping | Bright Data offers a collection of 600+ ready-made scrapers ... Oxylabs features an API Playground. |
| SE027 | Decodo | Residential Proxies: 115M+ Ethically-Sourced IPs in 195+ Locations | 115M+ ethically-sourced IPs in 195+ locations ... 99.92% success rate. |
| SE028 | Webshare | Proxy Pricing - Check Latest Proxy Price | 10 proxies free ... $0 /proxy. |
| SE029 | BestProxyCompare | 18 Best Proxy Providers in 2026, Ranked & Compared | We put 18 of the most popular providers through the same six-pillar scorecard. |
| SE030 | ScrapingBee | Best Rotating and Residential Proxies for Web Scraping in 2026 | If you need fine-grained control over raw proxies, other providers might be better, but ScrapingBee is recommended for teams ... wanting a managed experience. |
| SU001 | Oxylabs | Oxylabs home page | Trusted by 15,000+ clients globally. |
| SU002 | Oxylabs | Case Studies: Success Stories | From mid-market to enterprise, from SEO analytics to travel and hospitality, Oxylabs has made a difference. |
| SU003 | Oxylabs | Uninterrupted Retail Intelligence at Wiser Solutions | Oxylabs proxy offerings are a critical part of Wiser's workflows. |
| SU004 | Oxylabs | Epicup Provides Fresh SEO Data With Oxylabs Solutions | Oxylabs met all three, allowing us to offer real-time SEO data to our clients without breaking the bank. |
| SU005 | Oxylabs Blog | Case Study: Tackling Online Environmental Violations with Web Crawling Technology | Integrated into the department's daily operations since July 2024, the solution scans approximately 400 ads per week. |
| SU006 | Oxylabs | Oxylabs Reaches $3.6B Value After Agreeing to Its First Outside Investment | Trusted by over 350,000 tech teams worldwide. |
| SU007 | Oxylabs | Oxylabs AI Predictions 2026 | |
| SU008 | Webshare | Joining forces with Oxylabs | Now we serve over 15,000 customers from all over the globe who rely on Webshare everyday. |
| SU009 | Proxyway | Oxylabs Review & Performance Tests | Best enterprise proxy provider. |
| SU010 | Proxyway | The Best Residential Proxies of 2026 | Oxylabs ... Best premium residential proxy provider. |
| SU011 | Proxyway | 11 Best ISP (Static Residential) Proxies of 2026 | |
| SU012 | Proxyway | 7 Best US Proxy Providers of 2026 | |
| SU013 | Proxyway | The Best Rotating Proxies of 2026 | |
| SU014 | G2 | Oxylabs Reviews & Product Details | 4.5 ... 421 reviews. |
| SU015 | Capterra | Oxylabs Reviews 2026. Verified Reviews, Pros & Cons | Generally very good so far. Good support, stability of services and transparency on my spending. |
| SU016 | Trustpilot | Oxylabs is rated "Great" with 3.9 / 5 on Trustpilot | Be cautious with this company. Support is abysmal, and they do not refund. |
| SU017 | Gartner Peer Insights | Oxylabs Reviews & Ratings 2026 | Residential proxies are very easy to use ... valuable if you're running serious, high-volume scraping at an enterprise scale. Small operators may see better value elsewhere. |
| SU018 | Gartner Peer Insights | Oxylabs Reviews, Ratings & Features 2026 | |
| SU019 | Thunderbit | Oxylabs Review 2026: Pricing, Performance & Pitfalls | Enterprise users on Capterra give it a 4.7/5, while smaller operators on Reddit and Trustpilot tell a very different story. |
| SU020 | BestProxyCompare | 18 Best Proxy Providers in 2026, Ranked & Compared | If you're an enterprise with a big budget and specialist needs, shortlist Bright Data or Oxylabs. |
| SU021 | Decodo | Residential Proxies: 115M+ Ethically-Sourced IPs in 195+ Locations | 99.92% success rate ... Pay as you go. |
| SU022 | Webshare | Proxy Pricing - Check Latest Proxy Price | 10 proxies free. |
| SU023 | Oxylabs AI Studio | AI Studio – AI Web Scraping & Data Extraction Apps | Extract data from any webpage with AI. |
| SU024 | Oxylabs | Web Scraper API: Start Web Data Collection - Free Trial | Structured, ready-to-use data from multiple websites through one API. |
| SU025 | Oxylabs | Web Unblocker | Ensure high success rates with Oxylabs’ dynamic scraping strategy discovery system. |
| SR001 | Oxylabs Trust Center | Oxylabs Trust Center | We are excited to share that our Scraper API and Web Unblocker products have achieved SOC 2 Type 2 certification. |
| SR002 | Oxylabs | Risk and Legal Compliance | Our Trust Center is your hub for understanding how we protect data and maintain compliance. |
| SR003 | Oxylabs | KYC Policy | Every customer has to answer a KYC questionnaire to get access to our solutions. |
| SR004 | Oxylabs | General Conditions of oxylabs, UAB Services Agreement | Without our written consent you may not access the Services for purposes of monitoring their availability, performance or functionality or for any other competitive purposes. |
| SR005 | Oxylabs | Oxylabs Acceptable Use Policy | Automated data gathering (crawling, scraping, etc.) “as an activity itself” should not be considered illegal, however, there are certain methods and use cases when ... scraping might be considered ... in breach of relevant laws. |
| SR006 | Oxylabs | Oxylabs Data Processing Agreement | The Data Processor shall ... promptly notify the Data Controller in relation to any accidental or unauthorized access to Personal Data or any other security incidents. |
| SR007 | Cloudflare Docs | What is Pay Per Crawl? | Each time an AI crawler requests content, they either present payment intent ... or receive an HTTP 402 Payment Required response with pricing. |
| SR008 | EUR-Lex | Regulation (EU) 2024/1689 | Regulation (EU) 2024/1689. |
| SR009 | EUR-Lex | Regulation (EU) 2016/679 (GDPR) | Regulation (EU) 2016/679 ... on the protection of natural persons with regard to the processing of personal data. |
| SR010 | European Data Protection Board | Guidelines 03/2026 on web scraping in the context of generative AI | The GDPR applies to web scraping when it includes personal data processing operations. |
| SR011 | Oxylabs Blog | Introducing the Oxylabs Trust Center | The ISO 27001:2022 certificate is now available for review in our Trust Center. |
| SR012 | Farella Braun + Martel | Major Decision Affects Law of Scraping and Online Data Collection, Meta Platforms v. Bright Data | Meta argued that its terms prohibited users such as Bright Data from engaging in both logged-on and logged-off scraping. |
| SR013 | Quinn Emanuel | Bright Data: Questions Answered and Unanswered | The Bright Data ruling does not mean that scraping publicly available data is per se legal. |
| SR014 | Gartner Peer Insights | Oxylabs Reviews & Ratings 2026 | Valuable if you're running serious, high-volume scraping at an enterprise scale. Small operators may see better value in other providers. |
| SR015 | Trustpilot | Oxylabs is rated "Great" with 3.9 / 5 on Trustpilot | Be cautious with this company. Support is abysmal, and they do not refund. |
| SR016 | Thunderbit | Oxylabs Review 2026: Pricing, Performance & Pitfalls | Enterprise users on Capterra give it a 4.7/5, while smaller operators on Reddit and Trustpilot tell a very different story. |
| SR017 | Proxyway | Oxylabs Review & Performance Tests | Best enterprise proxy provider. |
| SR018 | Proxyway | The Best Residential Proxies of 2026 | Oxylabs ... Best premium residential proxy provider. |
| SR019 | BestProxyCompare | 18 Best Proxy Providers in 2026, Ranked & Compared | If you're an enterprise with a big budget and specialist needs, shortlist Bright Data or Oxylabs. |
| SR020 | Webshare | Proxy Pricing - Check Latest Proxy Price | 10 proxies free. |
| SR021 | Decodo | Residential Proxies: 115M+ Ethically-Sourced IPs in 195+ Locations | 99.92% success rate ... Pay as you go. |
| SR022 | TechCrunch | Cloudflare’s new policy pushes AI companies to pay for publishers’ content | Starting on September 15, 2026, Cloudflare’s default settings will block “mixed-use” crawlers from any pages that host ads. |
| SR023 | Oxylabs Blog | Oxylabs Reaches $3.6B Value After Agreeing to Its First Outside Investment | After a decade of growing profitably without external capital, Oxylabs has agreed to its first investment. |
| SR024 | Oxylabs Blog | Case Study: Tackling Online Environmental Violations with Web Crawling Technology | Integrated into the department's daily operations since July 2024, the solution scans approximately 400 ads per week. |
| SR025 | Oxylabs | Uninterrupted Retail Intelligence at Wiser Solutions | Oxylabs proxy offerings are a critical part of Wiser's workflows. |
| SR026 | Oxylabs | Epicup Provides Fresh SEO Data With Oxylabs Solutions | Epicup successfully built a robust, high-performing data gathering infrastructure. |
| SR027 | Webshare | Joining forces with Oxylabs | Now we serve over 15,000 customers from all over the globe who rely on Webshare everyday. |
| SR028 | Oxylabs | Oxylabs home page | Trusted by 15,000+ clients globally. |
| SR029 | Oxylabs | Web Unblocker | Ensure high success rates with Oxylabs’ dynamic scraping strategy discovery system. |
| SR030 | Oxylabs | Web Scraper API | Structured, ready-to-use data from multiple websites through one API. |
| SR031 | AI Act Explorer | The EU AI Act’s Transparency Rules: A Practical Guide to Article 50 | These obligations apply from 2 August 2026. |
| SR032 | ConductAtlas | Prohibition on Scraping and Automated Data Collection — LinkedIn User Agreement | Develop, support or use software, devices, scripts, robots or any other means ... to scrape or copy the Services. |
| SV001 | Oxylabs Blog | Oxylabs Reaches $3.6B Value After Agreeing to Its First Outside Investment | After a decade of growing profitably without external capital, Oxylabs has agreed to its first investment. |
| SV002 | Warburg Pincus | Oxylabs Announces $130 Million Investment from Warburg Pincus | Values Oxylabs at approximately $3.6 billion. |
| SV003 | The Next Web | Oxylabs raises $130M from Warburg Pincus at a $3.6bn valuation | Oxylabs says it now runs at $350m in annual recurring revenue. It counts more than 350,000 customers. |
| SV004 | SiliconANGLE | Web data scraping infrastructure startup Oxylabs reels in $130M in its first funding round | The company reckons it sees big demand for its web scraping services, counting more than 350,000 customers globally, which have helped it grow its annual recurring revenue to more than $350 million. |
| SV005 | EU-Startups | New Lithuanian unicorn: Oxylabs ends bootstrapped streak after securing €113.6 million at €3.1 billion valuation | The company currently serves more than 350,000 customers and generates €305.8 million ($350 million) in annual recurring revenue (ARR). |
| SV006 | CompaniesMarketCap | Similarweb (SMWB) - P/S ratio | P/S ratio as of July 2026 (TTM): 2.28 |
| SV007 | CompaniesMarketCap | ZoomInfo (GTM) - P/S ratio | P/S ratio as of July 2026 (TTM): 0.7347 |
| SV008 | CompaniesMarketCap | Cloudflare (NET) - P/S ratio | P/S ratio as of July 2026 (TTM): 45.7 |
| SV009 | CompaniesMarketCap | C3 AI (AI) - P/S ratio | P/S ratio as of July 2026 (TTM): 4.60 |
| SV010 | CompaniesMarketCap | Palantir (PLTR) - P/S ratio | P/S ratio as of July 2026 (TTM): 61.7 |
| SV011 | ZoomInfo | ZoomInfo Announces First Quarter 2026 Financial Results | GAAP Revenue of $310.2 million, an increase of 1.5% year-over-year. |
| SV012 | Similarweb | Similarweb Announces First Quarter 2026 Results | Total revenue was $73.9 million, an increase of 10% compared to $67.1 million. |
| SV013 | Cloudflare | Cloudflare Announces First Quarter 2026 Financial Results | Revenue: Total revenue of $639.8 million, representing an increase of 34% year-over-year. |
| SV014 | SEC / C3.ai | C3 AI Announces Preliminary Fourth Quarter and Full Fiscal Year 2026 Results | Total Revenue was $250.3 million. |
| SV015 | Palantir | Palantir Q1 2026 10-Q | a gross margin of 87% and 80%, respectively, or 88% and 82%, respectively, when excluding stock-based compensation. |
| SV016 | Similarweb | Similarweb Files 2025 Annual Report on Form 20-F | announced the filing of its annual report on Form 20-F for the fiscal year ended December 31, 2025, with the U.S. Securities and Exchange Commission (SEC) on March 2, 2026. |
| SV017 | ZoomInfo | ZoomInfo 10-Q filing page | Quarterly report which provides a continuing view of a company's financial position |
| SV018 | C3.ai | Quarterly Results SEC Filings | Quarterly Results SEC Filings |
| SV019 | Palantir | Palantir SEC filings | Filing date |
| SV020 | Gartner Peer Insights | Oxylabs Reviews & Ratings 2026 | valuable if you're running serious, high-volume scraping at an enterprise scale. Small operators may see better value elsewhere. |
| SV021 | Trustpilot | Oxylabs is rated "Great" with 3.9 / 5 on Trustpilot | Be cautious with this company. Support is abysmal, and they do not refund. |
| SV022 | Thunderbit | Oxylabs Review 2026: Pricing, Performance & Pitfalls | Enterprise users on Capterra give it a 4.7/5, while smaller operators on Reddit and Trustpilot tell a very different story. |
| SV023 | Proxyway | Oxylabs Review & Performance Tests | Best enterprise proxy provider. |
| SV024 | Proxyway | The Best Residential Proxies of 2026 | Oxylabs ... Best premium residential proxy provider. |
| SV025 | BestProxyCompare | 18 Best Proxy Providers in 2026, Ranked & Compared | If you're an enterprise with a big budget and specialist needs, shortlist Bright Data or Oxylabs. |
| SV026 | Oxylabs | Oxylabs home page | Trusted by 15,000+ clients globally. |
| SV027 | Oxylabs | Uninterrupted Retail Intelligence at Wiser Solutions | Oxylabs proxy offerings are a critical part of Wiser's workflows. |
| SV028 | Oxylabs | Epicup Provides Fresh SEO Data With Oxylabs Solutions | Epicup successfully built a robust, high-performing data gathering infrastructure. |
| SV029 | Oxylabs Blog | Case Study: Tackling Online Environmental Violations with Web Crawling Technology | Integrated into the department's daily operations since July 2024, the solution scans approximately 400 ads per week. |
| SV030 | TechCrunch | Cloudflare’s new policy pushes AI companies to pay for publishers’ content | Starting on September 15, 2026, Cloudflare’s default settings will block “mixed-use” crawlers from any pages that host ads. |