Linkerbot
在 $3B 估值下,出货和融资势头已有多方强证据支撑,但经审计财务、独立市场份额核验和可对账的融资历史仍缺位,尚不足以托住价格
Linkerbot 在灵巧手厂商中展现出少见且多方印证的出货规模和一线投资人背书,但审计财务、独立市占验证和融资轮次数据对账几乎缺位,意味着其 $3B 估值仍无法独立支撑。
封面要素
公司概况
Linkerbot 是一家总部位于北京海淀区的机器人硬件公司,约 2023 年成立,以 Linker Hand 品牌设计并量产高自由度灵巧机器人手,产品覆盖直驱、连杆和腱驱动架构,约 6–42 个自由度。公司称自己在全球高自由度灵巧手市场占有 80% 以上份额,截至 2025 年 12 月累计出货超过 10,000 台;截至 2026 年 5 月,公司正把月产能从约 5,000 台拉向 10,000 台目标,另有报道给出 2026 全年交付 50,000–100,000 台的目标。Linkerbot 在 2026 年 2 月完成约 RMB 1.5B 的 Series B,并在 2026 年 5 月前后以约 $3B 估值完成后续 ‘Series B+’,投资方包括 Ant Group、HongShan Group、Zhongguancun Science Park Fund、Bank of China Asset Management 和 Fosun Capital;公司还披露下一轮目标约 $6B。已点名客户覆盖工业制造(Samsung Electronics、Siemens)和研究机构(Stanford、MIT、Cambridge、University of Hong Kong)。所有已审阅来源均未披露收入、毛利率、现金状况、员工人数、董事会构成,也没有独立审计其市场份额主张。
- 成立时间
- 2023-01-01
- 创始人
- Alex Zhou (Zhou Yong)
- 创立地点
- Beijing, China
- 总部
- Beijing (Haidian District), China
- 产品
- Linker Hand 系列灵巧机器人手和末端执行器(13+ 个 SKU,覆盖 6–42 DoF,例如 O6、L20、L30)、‘Linker Craftsman’ 操作工作站,以及 LinkerSkillNet 操作技能软件生态和 Open TeleDex 遥操作系统。
- 客户
- 人形机器人集成商、工业 / 电子制造商(Samsung Electronics、Siemens)和研究机构(Stanford、MIT、Cambridge、University of Hong Kong);不同客户总数和收入集中度未披露。
- 商业模式
- 硬件优先,直接销售灵巧手 / 末端执行器产品;其上叠加一个正在成形的操作技能软件和遥操作平台,但后者是否单独贡献收入(如有)未披露。
- 阶段
- Series B+ private
- 融资情况
- 2025 年 11 月报道 Series A+(8 个月内第四次注资);2026 年 2 月报道约 RMB 1.5B 的 Series B;约 2026 年 5 月以约 $3B 估值完成 ‘Series B+’,并披露下一轮目标估值约 $6B;上市公司 Dongfang Precision(002611.SZ)披露直接持股约 4.36%。
执行摘要
主要优势
- 公司披露与多个独立追踪源互相印证,生产规模已跑出来:截至 2025 年 12 月累计出货 10,000+ 台,目标月产 10,000 台。
- 获得 Ant Group、HongShan Group、Zhongguancun Science Park Fund、Bank of China Asset Management、Fosun Capital 等一线机构支持,披露估值约 $3B。
- 开发者工具真正开放,GitHub 上有 ROS/ROS2 SDK、URDF 模型和模拟器,说明工程投入不只是营销口径。
主要风险
- 未披露审计财报、毛利率、现金状况或员工数;声称的 80%+ 市占率也没有独立审计。
- 关键人风险集中在唯一公开具名的高管身上,其身份 / 头衔仍有疑问,也未披露董事会或接班计划。
- 美国联邦采购禁令(GUARD Act、American Security Robotics Act)与中国 NDRC 对产能过剩的警示,带来双向监管和估值纪律风险。
未决问题
- 由独立第三方审计 Linkerbot 声称的 80%+ 高 DoF 市占率和单位出货量。
- 收入、毛利率、现金状况、烧钱速度和员工数披露。
- 对 2026 年 2 月 Series B 轮规模与 2026 年 5 月 Series B+ 估值做对账,以评估稀释和资本效率。
- 确认「Alex Zhou」与「Zhou Yong」是否为同一人,并披露董事会构成和接班安排。
目录
01公司概览
1.1 身份、总部和商业模式
Linkerbot 是一家总部位于北京海淀区的机器人硬件公司,为人形机器人、工业自动化和科研场景设计并量产高自由度(DoF)灵巧机器人手。公开报道将公司成立时间放在 2023 年;这与 2026 年 5 月 Reuters 来源报道把 Linkerbot 称为一家“成立两年的独角兽”的说法一致。核心产品是 Linker Hand 系列,覆盖直驱、连杆和腱驱动架构,约 6–42 个自由度,并由 CES 2026 展示的 ‘Linker Craftsman’ 工作站补充。公司自有站点地图至少列出 13 个不同产品页面(型号 L6、L10、L20、L20G、L30、L30Z、O6、O7、D、J、R、W 和 YCB),说明产品线分层且宽,而不是押注单一旗舰 SKU。Linkerbot 的商业模式以硬件为先:它制造并销售灵巧末端执行器给人形机器人集成商、工业买家和研究机构,再在硬件之上叠加软件 / 数据生态(LinkerSkillNet、Open TeleDex),而不是自己造整机机器人销售。截至 2026 年 5 月,公司阶段最适合描述为一家刚完成 ‘Series B+’ 的大型、未上市、风投资本支持的硬件独角兽;公司未公开上市,也未披露 IPO 时间表。[CO001, CO002, CO003, CO004, CO005, CO006]
| 指标 | 数值 / 状态 | 日期 | 置信度 | 缺口 |
|---|---|---|---|---|
| 总部 | 中国北京(海淀区) | 2026-07-04 | 高 | |
| 成立 | 2023(据报);准确注册成立日期未获独立确认 | 2023 | 中 | 通过一手登记备案确认准确注册成立日期 |
| 当前阶段 | 私营;Series B+ 已完成 | 2026-05 | 高 | |
| 最新披露估值 | 约 US$3B(Series B+) | 2026-05-04 | 高 | |
| 下一轮目标估值 | 约 US$6B(目标;时间 / 工具未披露) | 2026-05-04 | 中 | 确认时间,以及采用私募轮还是 IPO 工具 |
| 全球高自由度灵巧手市占率(声称) | 80%+ | 2026-05 | 中 | 缺少独立市占率审计 |
| 月产量(当前 -> 目标) | ~5,000 -> 10,000 台 / 月 | 2026-05 | 中 | 准确当前产能节奏未获独立确认 |
| 累计灵巧手出货量(截至 2025 年 12 月) | 10,000+ | 2025-12-18 | 中 | |
| 收入 / ARR | 未披露 | 低 | 未审阅公开财务报表或投资者材料 | |
| 人数 | 未披露 | 低 | 未找到公开人员规模披露 |
封面指标结合公司说法和 Reuters 来源通讯社报道;未获支持的数字(收入、人数)以 null 留空,并给出尽调路径。
[CO001, CO006, CO015, CO017, CO018, CO020]Linkerbot 把投资人资本转成手部制造规模,再服务工业 / 科研客户,并支撑软件层。
[CO005, CO016, CO025, CO026]1.2 领导层、治理和关键人依赖
Linkerbot 公开可见的领导层信息很薄,其中一个关键点仍未解开。2026 年 5 月 Reuters 来源通讯稿(由 BusinessWorld、Straits Times、Yahoo Finance 和 The Standard 再分发)把 ‘CEO Alex Zhou’ 称为讨论扩产和下一轮融资目标的高管。更早的 2025 年 11 月 China Biz Insider 中文报道则把 ‘Founder and CTO Zhou Yong’ 称为 Series A+ 融资阶段谈及订单量的人。已审阅来源没有明确说明 Alex Zhou 和 Zhou Yong 是否只是同一人的英文 / 中文名不同写法,也没有说明 2025–2026 两个报道窗口之间是否真的发生了 CTO 到 CEO 的头衔变化;本报告把它记录为未解决问题,而不是默认任一解释成立。除这位具名高管(或可能是同一人的两种称呼)外,已审阅来源没有披露独立董事会、外部董事、创始人股权分配或继任计划。已审阅报道中的产品战略、融资表态和制造扩产计划都通过同一位具名发言人传递;对于一家估值数十亿美元的公司,关键人依赖风险是实质性的。[CO008, CO009, CO010, CO011, CO012]
| 人员 / 实体 | 角色 | 背景或控制 | 创始人-市场匹配 / 职能覆盖 | 关键人依赖 |
|---|---|---|---|---|
| Alex Zhou | CEO(据 2026 年 5 月 Reuters 来源通讯社报道) | 作为高管被点名,引用于产能爬坡和市场战略表述 | 战略、爬坡和融资的唯一公开具名发言人 | 高——2026 年已审阅报道中唯一具名高管 |
| Zhou Yong | 'Founder and CTO'(据 2025 年 11 月 China Biz Insider 报道) | 在 Series A+ 阶段被引用讨论月订单量 | 早期报道中的技术 / 产品领导声音 | 是否与 Alex Zhou 为同一人仍不清楚,未解决 |
已审阅来源最多只点名两个人;他们是否是同一人在不同中英文命名和头衔下出现,仍未解决。
[CO008, CO009, CO010, CO012]1.3 融资历史、估值和投资方
Linkerbot 已披露的融资历史至少分成两个阶段。Series A+ 阶段,China Biz Insider 在 2025 年 11 月报道,公司刚完成 8 个月内第四次注资,投资方包括 Zhejiang Innovation Investment、CDH Investments 旗下 Baifu fund、Leju Robotics,以及上市战略投资方 Dongfang Precision 和 Aux,另有早期投资方 Ant Group 和 Sequoia China Seed Fund;该轮金额未披露。到 2026 年 4 月下旬 / 5 月上旬,Linkerbot 宣布 ‘Series B+’ 轮,估值约 US$3B,参投方包括 Ant Group、Sequoia 分拆后的 HongShan Group、国资背景的 Zhongguancun Science Park Fund、Bank of China Asset Management 和 Fosun Capital。同一公告中,公司称下一轮融资将寻求约 US$6B 估值,即 Series B+ 水平翻倍,但没有披露时间,也没有说明该轮会是私募融资还是公开上市。截至报告日,已审阅来源没有披露 Linkerbot 的累计总融资额;唯一可得的累计数字是 2025 年 12 月新闻稿中与出货里程碑一并披露的“自 2025 年 4 月以来融资 $150M”,该数字早于 $3B Series B+ 估值,也必然低估后者隐含的资本规模。已审阅来源没有披露任何二级转让、债务或授信安排。还需要注意的是,BusinessWorld、Straits Times、Yahoo Finance 和 The Standard 都佐证了 $3B / $6B 估值数字,但四者似乎都追溯到同一篇 Reuters 通讯稿加公司自有声明;因此佐证的分发面很广,原始报道链条却很窄。[CO013, CO014, CO015, CO016, CO017, CO018]
| 利益相关方 | 角色 | 控制或经济重要性 | 尽调问题 |
|---|---|---|---|
| Ant Group(Alibaba 关联方) | 投资者(Series A+ 和 B+ 阶段) | 战略财务投资者,触达中国金融科技 / 科技生态 | 确认董事席位或信息权 |
| HongShan Group(Sequoia China 分拆) | 投资者(Series B+) | 中国大型创投机构;释放后期机构信心信号 | 确认持股比例 |
| Zhongguancun Science Park Fund | 投资者(Series B+) | 国资背景基金;释放地方政府产业政策支持信号 | 确认治理权利及任何战略条件 |
| Bank of China Asset Management(投资方) | 投资者(Series B+) | 国资背景金融机构;带来资产负债表级资本 | 确认工具(股权还是结构化) |
| Fosun Capital | 投资者(Series B+) | 私营综合集团投资平台 | 确认持股规模及与公司的既有关系 |
| Zhejiang Innovation Investment、CDH Baifu 与 Leju Robotics | 投资者(Series A+ 阶段,2025 年 11 月) | 较早期资本及潜在战略 / 产业伙伴 | 确认 Series B+ 后是否仍持股 |
| Dongfang Precision / Aux(上市公司) | 战略投资者(Series A+ 阶段) | 上市公司战略投资者,可能有供应链或渠道联系 | 确认是否存在与投资绑定的商业协议 |
投资者名单来自公司说法,以及 Reuters / China Biz Insider 对两个融资阶段的报道;准确持股比例未披露。
[CO013, CO014, CO016]约三年内,Linkerbot 从创立走到 $3B「Series B+」估值,并首次亮相 CES。
[CO002, CO013, CO015, CO017, CO020, CO022]1.4 封面指标:规模、市场份额、客户和披露缺口
Linkerbot 最常被引用的封面指标,是其声称在高自由度机器人手全球市场的份额:公司称自己占有 80% 以上份额,并且是唯一一家每月量产超过 1,000 只高自由度灵巧手的制造商;这一说法同时出现在公司新闻稿和 Reuters 来源报道中。独立行业跟踪方 TechBuzz China 与中文市场分析 Gongboshi Robot 都把 Linkerbot 描述为截至 2026 年上半年拥有 80%+ 份额的品类领导者,在公司自述之外提供了部分第三方佐证,但两者都没有披露底层方法论。产能方面,Linkerbot 在 2026 年 5 月称其正把月产量从接近 5,000 台提升到“很快”达到计划中的 10,000 台;此前公司已披露截至 2025 年 12 月累计出货超过 10,000 只手,单月最高超过 4,000 台。已披露的具名客户仅限工业 / 生产用途的 Samsung Electronics 和 Siemens,以及被描述为包括 Stanford 和 MIT 在内的科研用途;没有来源给出合同金额、按客户拆分的台数或续约信息。本章审阅的所有来源均未披露收入、ARR、毛利率和员工人数,Linkerbot 的 HKTDC 国际采购店铺页面还返回“页面未找到”,这是一个小但具体的渠道可见性缺口。另一个需要单列的背景信号是,中国国家发展和改革委员会在 2026 年公开指出中国已有 150 多家人形机器人公司,并警示过度产品重复和投机泡沫;这是行业层面的负面信号,关系到外界应如何解读单家公司激进的估值轨迹。[CO020, CO021, CO022, CO023, CO024, CO025]
| 日期 | 事件 | 类型 | 金额 / 估值 / 状态 | 参与方 | 影响 |
|---|---|---|---|---|---|
| 2023 | 公司成立 / 实体在北京注册 | 创立 | N/A | 创始团队 | 为快速爬坡硬件业务确立起点 |
| 2025-11-25 | Series A+ 融资据报完成;8 个月内第 4 次注资 | 融资 | 未披露(轮次规模) | Zhejiang Innovation Investment、CDH Baifu、Leju Robotics、Dongfang Precision、Aux、Ant Group 与 Sequoia China Seed Fund | 在广泛公开可见前,释放机构快速反复融资信号 |
| 2025-12-18 | 第 10,000 只灵巧手出货;披露自 2025 年 4 月以来已融资 $150M | 规模 | $150M 累计(截至该日期) | Linkerbot | 首个公开累计出货和融资里程碑 |
| 2026-01-08 | CES 2026 首发完整 Linker Hand 产品线和「Linker Craftsman」工作站 | 产品 | N/A | Linkerbot | 公开国际产品展示,扩大品牌可见度 |
| 2026-04-30 至 2026-05-04 | 「Series B+」轮以约 $3B 估值完成;披露 $6B 下一轮目标 | 融资 | $3B 估值(当前);$6B(目标) | Ant Group、HongShan、Zhongguancun Science Park Fund、Bank of China Asset Management 与 Fosun Capital | 标志进入独角兽以上状态,并为下一轮融资设下公开标尺 |
| 2026-05 | Reuters 访谈:披露产能从约 5,000 台 / 月爬坡到 10,000 台 / 月 | 规模 | 5,000 -> 10,000 台 / 月 | CEO Alex Zhou | 显示继续制造爬坡的雄心 |
| 2026(未注明日期) | 国家发改委警示人形机器人行业产能过剩 / 泡沫风险 | 监管 | 提及 150+ 家公司 | 中国国家发改委 | 与行业估值纪律有关的负面宏观信号 |
| 2026(持续) | HKTDC 供应商店铺页面返回 not-found 结果 | 不利 | N/A | HKTDC / Linkerbot | 某个国际贸易目录中的轻微披露 / 渠道可见性缺口 |
时间线结合公司新闻稿、Reuters 来源通讯社报道,以及一个不利宏观 / 监管数据点;若来源只给出月份,若干日期采用近似区间。
[CO002, CO013, CO015, CO020, CO022, CO023]公开记录在估值、市占率和出货节奏上最强,在财务报表上最弱。
[CO015, CO017, CO020, CO021]1.5 图表
02市场分析
2.1 市场边界、相邻市场和现状替代方案
Linkerbot 真正竞争的市场,比“人形机器人”整体要窄:它是具身 AI 硬件栈中专注灵巧末端执行器的子赛道,即具备类人自由度的机器人手和夹爪,再加相邻的操作软件层(LinkerSkillNet、Open TeleDex)。完整人形机器人的更大 BOM 并不计入这块可服务支出:底盘、运动系统、电池和通用 AI 算力都由其他供应商提供,或由机器人集成商自研。相邻市场包括 Unitree、Tesla Optimus 等完整人形机器人厂商,它们目前自研手部而不是向供应商采购;也包括工业自动化夹爪老牌厂商,它们可能逐步增加自由度,并从低端挤入 Linkerbot 的品类。买家如果不需要完整类人灵巧度,最直接的现状替代品就是简单的 2–3 指平行夹爪(如 RightHand Robotics 等老牌厂商所售),或者继续用人工完成尚不具备自动化成本效益的精细操作任务。这样界定市场很重要,因为整个人形机器人 TAM 数字无论多大,都会系统性高估 Linkerbot 自己能捕获的部分。[CM001, CM002, CM003, CM004]
| 细分市场 / 类别 | 纳入支出 | 排除支出 | 买方 / 付款方 | 与 Linkerbot 的关系 |
|---|---|---|---|---|
| 高自由度灵巧机器人手 / 末端执行器 | 手部 / 夹爪硬件、触觉传感模块、操作技能软件(如 LinkerSkillNet) | 完整机器人底盘、运动系统、电池、通用 AI 计算 | 机器人集成商、工业买家、研究实验室 | 直接可服务市场——Linkerbot 的核心产品 |
| 完整人形机器人系统 | 包含手部的完整双足 / 人形机器人 BOM | N/A(上位集合) | 机器人集成商的终端客户(工厂、物流、消费者) | 相邻——更大的市场,包含但不限于手部 |
| 工业平行夹爪 / 简单夹爪 | 用于拾放的低自由度夹爪 | 高自由度灵巧操作 | 工业自动化买家 | 不需要类人灵巧度时的现状替代方案 |
| 假肢 / 康复机器人 | 医疗用途辅助灵巧手 | 工业 / 人形机器人用例 | 医院、患者、保险方 | 行业跟踪提及的相邻新兴用例,尚未披露为 Linkerbot 细分市场 |
| 遥操作抓取 / 操作系统 | 人在环路遥操作硬件 / 软件 | 全自主操作 | 试点部署的机器人集成商 | 相邻 / 互补——Linkerbot 的 Open TeleDex 位于这一边界 |
市场边界专门划在灵巧手 / 末端执行器;完整人形机器人 TAM 数字只作背景,并不直接等同于 Linkerbot 的可服务支出。
[CM001, CM002, CM003, CM004]2.2 从多个相互冲突的视角测算 TAM/SAM/SOM
这个市场没有一个可信的单一 TAM 数字。2026 年整个人形机器人市场规模估计,从 Grand View Research 的约 $4.2B、MarketsandMarkets 的 $5.41B,到 Future Market Insights 的 $10.69B、Global Market Insights 的 $10.9B 不等;同一年份口径相差超过 2 倍,且已检索到的出版商页面摘录都没有完整披露底层方法论。Goldman Sachs Research 则把机会描述成长期情景区间:到 2035 年,基准情景接近 $38B;若可靠性和成本障碍下降快于预期,“blue sky” 情景可达 $154B。这些整体市场数字都没有单独拆出 Linkerbot 实际销售进入的灵巧手子赛道。最接近的代理值来自 China Biz Insider 的估计:中国本土灵巧手市场到 2031 年将超过 $3.7B;同时,行业跟踪方观察到中国供应商已提供全球 70–80% 的商用高自由度机器人手,单价也从超过 $100,000 降至低于 $2,000。即使这个更窄的数字也只是单篇文章引用,而非经审计的市场研究;已审阅来源也没有给出 Linkerbot 专属的可服务可获得市场(SOM)数字。公司自称 80%+ 份额所对应的品类边界,是其自己定义的“全球高自由度灵巧手市场”,不是任何具名第三方测算研究。[CM005, CM006, CM007, CM008, CM009, CM010]
| 发布方 | 年份 | 地域 | 数值 | CAGR | 方法论(如披露) | 置信度 | 局限 |
|---|---|---|---|---|---|---|---|
| MarketsandMarkets | 2026 | 全球 | $5.41B(完整人形机器人市场) | 28.1% (2026-2035) | 本次审阅摘录未完整披露 | 中 | 整体市场数字,不是灵巧手专属 |
| Global Market Insights 研究 | 2026 | 全球 | $10.9B(完整人形机器人市场) | 37.6% | 本次审阅摘录未完整披露 | 中 | 是同年 MarketsandMarkets 的 2 倍以上;方法论未勾稽 |
| Future Market Insights 研究 | 2026 | 全球 | $10.69B(完整人形机器人市场) | 37.0% (2026-2036) | 本次审阅摘录未完整披露 | 中 | 整体市场数字,不是灵巧手专属 |
| Grand View Research | 2026 | 全球 | $4.2B(完整人形机器人市场,归档估算) | 38.2% (2026-2033) | 未完整披露;归档快照 | 中 | 快照可能早于报告日期;使用归档副本 |
| Goldman Sachs Research | 2035 | 全球 | $38B 基准情景 / 最高 $154B 蓝天情景 | 未按 CAGR 表述 | 基于情景,不是点估计 | 高 | 长周期情景区间,不是 2026 年点估计 |
| China Biz Insider(灵巧手 SAM 代理) | 2031 | 中国 | >$3.7B(仅中国灵巧手市场) | 41.6%(隐含) | 未独立审计;单篇文章引用 | 低 | 与 Linkerbot 关系最窄、最直接的视角,但独立验证最弱 |
完整人形机器人 TAM 数字仅作背景;中国灵巧手数字是最接近的 SAM 代理,但独立印证最弱。
[CM005, CM006, CM007, CM008, CM011]2026 年整个人形机器人 TAM 估算远高于 Linkerbot 实际竞争的更窄灵巧手 SAM 代理。
各层混合了不同发布方、年份和地域;应视为说明性漏斗,而非单一经审计的 SAM/SOM 瀑布图。
[CM001, CM008, CM009, CM011]已发布的 2026 年整个人形机器人 TAM 估算约为 $4.2B 至 $10.9B,取决于发布方。
每家发布方只给出一个 2026 年点估计;这里的「区间」是发布方之间的差异,不是某一家发布方的低 / 基准 / 高情景。单位统一为十亿美元。
[CM006, CM011, CM026]2.3 买方、用户、付款方分层和采用路径
Linkerbot 这类灵巧手的买方至少分成三类,预算归属不同。工业和电子制造买家——包括 Linkerbot 自己披露的客户 Samsung Electronics 和 Siemens——通常从自动化 / 资本开支预算中采购,买方、用户和付款方属于同一企业职能,采用触发点是相对于既有夹爪或人工流程的可靠性与成本平价。Unitree 等人形机器人集成商则构成独立的“买还是自研”客群:它们可以向 Linkerbot 这样的供应商采购手部,也可以自己设计,因此采用触发点不是简单采购决策,而是成本和上市时间比较。研究机构(Linkerbot 称使用方包括 Stanford 和 MIT)通过科研经费和实验室预算采购,而不是企业采购;采用触发点是论文发表和演示价值,而非商业 ROI,因此这是低收入但高可见度的渠道。一项 Morgan Stanley 引用的 2026 年调查发现,许多中国企业买家希望引入人形 / 灵巧自动化,但尚不认为现有产品可靠到足以全量投入生产;这说明行业典型采用漏斗——认知、试点、认证、生产集成、复购——仍集中在早期阶段,Linkerbot 也不例外,尽管没有来源披露 Linkerbot 自身的漏斗转化数据。[CM012, CM013, CM014, CM015, CM016, CM031]
| 细分市场 | 买方 | 用户 | 付款方 | 工作流 | 预算所有者 | 采用触发点 |
|---|---|---|---|---|---|---|
| 工业 / 电子制造 | 企业采购(如公司说法中的 Samsung、Siemens) | 产线技术员 / 自动化工程师 | 同买方 | 装配、精密处理、QA 任务 | 资本开支 / 自动化预算 | 可靠性 + 成本与现有夹爪或人工流程持平 |
| 人形机器人集成商 | 机器人 OEM(如 Unitree、其他中国 / 全球人形机器人厂商) | OEM 工程团队 | 与买方相同,或由最终客户通过机器人打包价格承担 | 手部模块集成进完整人形机器人产品 | 研发 / BOM 预算 | 外采与自研的成本和上市速度对比 |
| 科研机构 | 高校实验室(公司称 Stanford、MIT 在用户之列) | 研究生 / PI | 课题经费 | 操作研究、数据集采集 | 课题 / 实验室预算 | 论文与演示价值,而非商业 ROI |
| 假肢 / 康复(新兴) | 医院、设备厂商 | 患者 | 保险方 / 患者 / 设备厂商 | 日常生活辅助任务 | 医疗资本开支或报销 | 监管审批与临床验证 |
细分市场图把 Linkerbot 自己披露的客户,与行业报道推断出的细分结构合并;没有来源披露各细分市场的收入占比。
[CM012, CM013, CM014, CM015, CM016]三个买方细分在预算归属和采用触发点上差异很大。
[CM012, CM013, CM014, CM033]已审阅的 2026 年证据大多显示,买方仍处在漏斗前段(试点 / 评估),而非生产复购阶段。
漏斗比例为说明性估算,来自定性调研评论(引用 Morgan Stanley),不是 Linkerbot 自身披露的数字漏斗。
[CM016, CM025, CM031]2.4 增长驱动因素和采用约束
几股力量正在推动采用。中国“十五五”规划(2026–2030)把 AI 机器人提升为核心国家战略重点,为本土手部厂商提供持续政策顺风。单只手价格已大幅下降:同代中国制造灵巧手从超过 $100,000 降到低于 $2,000,一些竞争对手还公开预计几年内可降到 $100 以下;买方池因此有望从资金充足的机器人实验室扩展到更广的客户。竞争性人形机器人厂商在 2026 年的公开演示,也抬高了全行业投资人和买方关注度。反向约束同样真实且重要。中国国家发展和改革委员会已公开警示,国内已有 150 多家人形机器人公司,并提醒避免产品重复和投机泡沫;这意味着后续可能出现政策收紧或投资人转谨慎。需求端,美国国会议员在 2026 年 3 月提出 GUARD Act 和 American Security Robotics Act,二者都旨在禁止美国联邦政府采购和使用中国制造的人形及地面机器人;这直接约束包括灵巧手在内的中国制造部件进入美国公共部门需求。更广泛的美国半导体和先进计算出口管制,也可能进一步限制销往或经由美国关联供应链的高端手部电子件。最后,国际安全标准(ISO 10218、ISO/TS 15066)和仍处早期的可靠性曲线——见 Morgan Stanley 引用的企业调查——意味着许多工业买家仍停留在试点,而不是承诺生产规模、多台套订单。[CM017, CM018, CM019, CM020, CM021, CM022]
| 驱动因素 / 约束 | 方向 | 时点 | 影响 | 尽调问题 |
|---|---|---|---|---|
| 中国“十五五”规划把 AI / 机器人提升为核心战略 | 驱动 | 2026-2030 | 国家层面对本土灵巧手厂商的需求和资本支持更可持续 | 确认 Linkerbot 是否直接享受补贴或采购偏好 |
| 单只灵巧手价格下降(低于 $2,000;按竞争对手说法,数年内趋向低于 $100) | 驱动 | 持续 | 把买方池从资金充足的实验室扩大到主流工业买家 | 确认 Linkerbot 自身价格点和成本曲线 |
| 竞争对手人形机器人厂商公开演示,抬升行业能见度 | 驱动 | 2026(持续) | 抬高整个行业的投资人与买方关注度,间接利好灵巧手供应商 | 评估关注度是否转化为 Linkerbot 自身订单增长 |
| NDRC 泡沫 / 产能过剩警告(中国 150+ 家人形机器人公司) | 约束 | 2026(持续) | 预示后续可能出现政策收紧或投资人趋谨慎 | 监测具体政策动作(补贴上限、许可等),不要只看表态 |
| 美国 GUARD Act / American Security Robotics Act 联邦采购禁令 | 约束 | 自 2026 年起 | 切断美国联邦 / 公共部门对中国制造机器人及其可能部件的需求 | 确认销往美国集成机器人中的 Linkerbot 灵巧手是否落入适用范围 |
| 美国半导体 / 先进计算出口管制 | 约束 | 持续 | 可能限制销售给美国关联供应链或经其销售的高端灵巧手内部电子与计算部件 | 确认 Linkerbot 的半导体采购,以及是否暴露于美国成分 |
| ISO 10218 / ISO-TS 15066 安全与认证标准 | 约束 | 持续 | 大型工业买家通常要求先拿到认证再投入量产部署 | 确认 Linkerbot 针对这些标准的认证状态 |
| Morgan Stanley 引用的企业调查提示可靠性 / 量产就绪度缺口 | 约束 | 2026(当前) | 即便有预算和兴趣,也会拖慢企业采用 | 索取已量产部署(非试点)的参考客户 |
驱动因素与约束来自公司、政策、分析师等多类来源;其中若干约束是行业层面的,并非 Linkerbot 独有。
[CM017, CM018, CM019, CM020, CM021, CM022]2.5 图表
03竞争对手
3.1 竞争格局:同类厂商、老牌厂商、相邻玩家和替代品
Linkerbot 的竞争集合横跨三类。第一类是与 Linkerbot 一样把灵巧手作为独立产品销售的直接灵巧手专家:Inspire Robots(因时机器人)、Shadow Robot(英国),以及已进入该品类的零部件制造商 Zhaowei Mechatronics。第二类,也可能是战略上更重要的一类,是自研并制造手部、而不是向独立供应商采购的全栈人形机器人厂商——Unitree、AgiBot、Fourier Intelligence、UBTech 和 Tesla 都在其中。它们既可能成为潜在客户(如果未来选择采购而非自研),也会构成永久竞争威胁(因为它们可能出售富余手部产能,或像 AgiBot 已开始做的那样拆出手部业务)。第三类是非中国人形机器人厂商——Figure AI、Apptronik 和 1X Technologies——目前在内部自研手部,但也是未来美国政策行动最可能重塑全球采购模式的客群。专用灵巧手之外,任何不需要完整类人灵巧度的买家,其现状替代方案是简单平行夹爪(如 RightHand Robotics 等老牌厂商所售),或继续依靠人工。低成本、实验室导向的相邻产品——Wonik Robotics 的 Allegro Hand 和开源 LEAP Hand——服务研究买家而非工业客户,不太可能直接替代 Linkerbot 的工业客户基础。[CP001, CP002, CP003, CP004, CP029, CP031]
| 竞争对手 | 类别 | 规模 / 融资 | 目标细分市场 | 差异化 | 局限 |
|---|---|---|---|---|---|
| Unitree Robotics | 全栈人形机器人厂商(自研灵巧手) | 估值 ~$1.6B(2025 年 Series C 轮);已提交 IPO,据报募资最高 $580M | 工业、科研、消费级演示 | 整机 + 灵巧手一体化;Dex5-1/Dex3-1/Dex1-1 分层手部产品族 | 灵巧手不单独销售;没有独立手部定价 |
| AgiBot(OmniHand) | 全栈人形机器人厂商,转向单独销售灵巧手 | 审阅来源未披露 | 工业 + 部件转售 | 10-DoF OmniHand 以 ~$1,350 单独销售 | 相近价格点下,自由度低于 Linkerbot 旗舰层级 |
| Fourier Intelligence | 全栈人形机器人厂商(GR-1/GR-2) | 审阅来源未披露 | 科研、康复、服务机器人 | 有康复机器人积累,并向人形机器人拓展 | 买方结构不同于 Linkerbot 的工业 / 精细作业重心 |
| UBTech Robotics | 全栈人形机器人厂商(Walker S2) | 披露 Walker S2 订单 >RMB 800M | 工业规模部署 | 上市公司,披露了商业订单积压 | 手部规格未从机器人打包方案中单独披露 |
| Shadow Robot(英国) | 直接灵巧手专精厂商(高端) | 未披露;单价 >$140K | 科研、高端机器人实验室 | 24 DoF、100+ 传感器——公认技术标杆 | 在这一价格点,规模化生产已经停滞 |
| Zhaowei Mechatronics | 直接灵巧手新进入者(部件制造背景) | 审阅来源未披露 | 工业、部件供应 | DM17(17 个主动 DoF)和 LM06(150N 握力)产品线 | 新进入者;在灵巧手领域的品牌认知低于 Linkerbot |
| Tesla(Optimus) | 全栈人形机器人厂商(自研手部) | 未作为独立手部业务披露 | 通过 Tesla 生态面向消费 / 工业 | 估计 Optimus 约一半工程投入用在手部;机器人量产成本目标 $20K-$25K / 台 | 手部规格来自分析师 / 拆解估算,并非公司披露 |
| Figure AI(美国) | 全栈人形机器人厂商(自研手部) | 投后估值 $39B,Series C 轮融资 >$1B | 美国工业 / 物流 | 资金最充足的西方人形机器人竞争者 | 未披露独立手部定价或中国市场存在 |
| Apptronik(美国) | 全栈人形机器人厂商(自研手部) | $520M 融资,估值 $5B | 美国工业,抢在 Tesla 之前进入市场的定位 | 明确策略是先于 Tesla 进入市场 | 资本基础小于 Figure AI,也小于 Linkerbot 目标估值 |
| 1X Technologies(美国 / 挪威) | 全栈人形机器人厂商(自研手部) | $100M Series B 轮 | 消费 / 家庭机器人重心 | 相对工业重心,家庭机器人定位有差异 | 披露融资在具名西方同业中最少 |
| X Square Robot | 相邻具身 AI / 人形机器人厂商 | 连续 4 轮后估值达 $2.8B | 家用 / 消费级操作任务 | 家庭任务重心鲜明,不同于 Linkerbot 的精细作业 / 工业重心 | 终端市场不同,降低了与 Linkerbot 的直接竞争重叠 |
竞争对手集合覆盖直接灵巧手专精厂商、把手部自研的全栈人形机器人厂商,以及相邻具身 AI 厂商;融资数字来自各公司 / 新闻来源分别披露,并非同一套统一数据集。
[CP004, CP005, CP006, CP007, CP008, CP009]3.2 竞争对手画像:规模、融资和战略方向
Unitree 在 2026 年 3 月 IPO 招股书中披露了三档自研灵巧手:20-DoF Dex5-1、7-DoF Dex3-1 和 Dex1-1 夹爪;在 2025 年中 Series C 后估值约 $1.6B,之后在人形机器人销售开始超过传统四足机器人业务时提交 IPO,报道称拟募资最高 $580M。AgiBot 已开始独立销售 OmniHand 系列,一款 10-DoF 型号定价约 $1,350,直接打破过去只把手部与整机捆绑销售的做法。Shadow Robot 仍是公认技术标杆,具备 24 DoF 和 100+ 个传感器,但单手价格超过 $140,000,量产已停滞——说明原始技术领先没有转化为规模。Zhaowei Mechatronics 从零部件制造背景切入,提供 DM17(17 个主动 DoF)和 LM06(150N 抓握力)产品线。在全栈西方竞争者中,Figure AI 资本领先,Series C 融资超过 $1B 后投后估值 $39B,高于 Apptronik(融资 $520M、估值 $5B)和融资规模更小的 1X Technologies($100M Series B)。中国具身 AI 竞争者 X Square Robot 连续四轮融资后估值超过 $2.8B,但明确瞄准家庭 / 消费操作任务,而非 Linkerbot 所称的工业 / 工匠式重点,因此直接重叠较小。UBTech 的 Walker S2 人形机器人已开始量产,披露订单超过 8 亿元;Galbot(融资 RMB 2.5B)和 Robotera(融资超过 $200M)也显示,新资本仍在进入 Linkerbot 灵巧手最终服务或竞争的更广泛中国人形机器人生态。[CP005, CP006, CP007, CP008, CP009, CP010]
| 采购标准 | Linkerbot | Unitree | AgiBot OmniHand | Shadow Robot | Zhaowei Mechatronics |
|---|---|---|---|---|---|
| 自由度(旗舰层级) | 最高 42 DoF(Linker Hand 系列) | 20 DoF(Dex5-1) | 10 DoF(独立型号) | 24 DoF | 17 个主动 DoF(DM17) |
| 触觉感知 | 多模态“感知皮肤”(公司声称) | 94 个触觉传感器(Dex5-1) | 审阅来源未披露 | 100+ 传感器 | 审阅来源未披露 |
| 单独销售单价(已披露) | ~¥6,666 (~$920),O6 型号 | 不单独销售 | ~¥9,800 (~$1,350) | >¥1,000,000 (~$140,000) | 审阅来源未披露 |
| 已披露月产能规模 | ~5,000 -> 10,000 只 / 月(目标) | 手部层面未披露 | 审阅来源未披露 | 据 techbuzzchina,停滞 / 低产量 | 审阅来源未披露 |
| 作为独立部件销售,还是仅打包 | 独立销售(专注灵巧手公司) | 与机器人打包 | 独立销售(截至 2026 年定价动作) | 独立销售(科研 / 实验室销售) | 独立销售(部件供应商模式) |
无支撑的单元格标为“审阅来源未披露”,而非估算;规格来自各公司披露或独立媒体 / 分析师报道,测试方法各不相同。
[CP005, CP007, CP008, CP009, CP017, CP033]相较于 Shadow Robot(高 DoF、价格很高)和 AgiBot / Zhaowei(中等 DoF、低价),Linkerbot 落在高 DoF、低价象限。
坐标轴为顺序尺度(有证据支撑的相对定位:x=技术 DoF / 能力,y=相对价格层级),不是精确数值刻度,因为并非每个竞争对手都披露标准化规格或价格。
[CP005, CP007, CP008, CP009, CP017]3.3 定价、GTM 与信任 / 监管对比
已披露价格中,Linkerbot O6 型号约 $920,低于 AgiBot 独立销售的 10-DoF 手(约 $1,350),也远低于 Shadow Robot 的技术标杆手(>$140,000)和行业跟踪方引用的海外通用品类可比价格($150,000–$220,000)。Unitree、Fourier Intelligence 和 UBTech 都没有披露独立手部定价,因为它们只把手部捆进完整人形机器人销售;这意味着在具名全栈 / 技术标杆同业中,真正可做同口径价格比较的只有 AgiBot 和 Shadow Robot。GTM 方面,Linkerbot 已披露渠道是国际新闻和展会可见度(CES 2026、PR Newswire 分发)叠加对具名工业和研究账户的直销;已审阅来源没有披露正式经销商或分销商网络,这不同于全栈厂商通过自身机器人销售组织触达终端客户。信任与监管姿态方面,已审阅来源没有披露任何具名竞争者灵巧手产品的正式安全或质量认证——Linkerbot 自身也没有——除市场分析章节讨论的行业层面 ISO 工业机器人安全标准外。因此,认证状态是整个竞争集合的开放尽调项,而不是 Linkerbot 独有弱点。[CP017, CP018, CP019, CP020, CP030]
| 竞争对手 / 型号 | 价格 / 单位 | 包含能力 | 折扣或未知项 | 影响 |
|---|---|---|---|---|
| Linkerbot O6 | ~¥6,666 (~$920) | 6 DoF,重量 370g,50kg 握持载荷(公司声称) | 仅标价;批量折扣条款未披露 | 相较多数具名对手,Linkerbot 卡在大众市场价格点 |
| AgiBot OmniHand(10-DoF) | ~¥9,800 (~$1,350) | 10 DoF 独立灵巧手 | 打包与单独销售的折扣条款未披露 | AgiBot 价格低于更高 DoF 竞争对手,但高于 Linkerbot O6 价格点 |
| Shadow Robot Dexterous Hand | >¥1,000,000 (~$140,000) | 24 DoF,100+ 传感器 | 定制 / 科研定价可能因配置而异 | 凸显技术标杆与大众市场中国灵巧手之间的价格差 |
| 海外同类通用灵巧手 | ¥1.1M-1.6M (~$150K-$220K) | 按厂商而异 | 来自单一行业追踪器的汇总估算,并非厂商特定数据 | 刻画中国在这一品类中的成本优势量级 |
所有价格均为公开披露或第三方追踪器估算;均不代表经审计、发票级的实际成交价。
[CP017, CP008]3.4 切换成本、锁定、分销权力和护城河耐久性
一旦机器人集成商围绕某个供应商的手部工程化了控制栈、安装结构、线束和操作技能数据,切换成本很可能不低;这会为被设计进某一机器人平台的供应商创造真实但尚未量化的锁定。分销权力则更多掌握在全栈人形机器人厂商(Unitree、AgiBot、Figure AI、UBTech)手中,因为它们控制终端客户关系;Linkerbot 这样的独立手部供应商,要么被设计进别人的机器人,要么直接卖给研究 / 工业账户,分销位置结构性弱于集成商本身。Linkerbot 自称的护城河基于生产规模和成本领先——80%+ 自称份额、低于 $1,000 的定价、1,000+ 台 / 月的量产下限——而非单一技术突破;Shadow Robot 已在原始 DoF 上保持更高技术标杆,只是成本高得多。这个规模型护城河面临两个具体威胁:商品化风险,AgiBot 和 Zhaowei 以相近或更低价位进入已说明这一点;以及围绕自报规模主张的可信度风险,Humanoids Daily 报道的 Unitree 与 AgiBot 关于全球出货领先权公开争议就是例证。中国国家发展和改革委员会 2026 年警示国内已有 150 多家人形机器人公司,这是另一个行业层面的负面信号:产能过剩或整合可能压缩整个竞争集合的利润率和估值,Linkerbot 也在其中。已审阅来源没有记录 Linkerbot 对任何具名对手的具体赢单或败单,因此这里的护城河耐久性很大程度上来自推断,而非直接证据。[CP021, CP022, CP023, CP024, CP025, CP026]
| 护城河主张 | 威胁 | 严重性 | 缓解 / 尽调问题 |
|---|---|---|---|
| “全球高 DoF 灵巧手市场”80%+ 份额 | 自定义品类,没有独立审计;Unitree-AgiBot 出货量争议说明同业会挑战类似主张 | 重大 | 在把该数字用于估值前,委托独立审计出货量 / 市占率 |
| 成本 / 规模领先(低于 $1,000 定价,1,000+ 只 / 月) | AgiBot 和 Zhaowei 以相近或更低价格点进入独立灵巧手销售,可能把品类商品化 | 重大 | 按季度跟踪竞争对手价格动作;评估 Linkerbot 成本曲线相对对手的走势 |
| 宽产品线(13+ SKU,6-42 DoF) | Unitree、AgiBot、Fourier、UBTech 等全栈集成商自研手部,长期会压缩可触达买方池 | 重大 | 监测主要集成商是否继续购买第三方灵巧手,还是完全转向自研 |
| 软件 / 平台层(LinkerSkillNet、Open TeleDex) | 硬件销售之外,软件层没有披露采用指标;可能只是营销叙事,而非耐久数据护城河 | 较小 | 索取软件平台使用 / 采用数据,并与硬件销量拆开 |
| 行业整体产能过剩风险(NDRC 警告,中国 150+ 家人形机器人公司) | 行业大范围整合或融资收缩可能压低所有竞争者的利润率 / 估值,包括 Linkerbot | 重大 | 监测 NDRC / 政策信号,以及竞争对手倒闭或整合事件 |
严重性为定性判断,基于已审阅证据权重,并非量化风险模型。
[CP024, CP025, CP026, CP027, CP004, CP034]Linkerbot 在已披露 DoF 范围和价格可及性上领先;Shadow Robot 在原始技术基准上领先;多家竞争对手未披露独立销售规格。
[CP005, CP007, CP008, CP024]Linkerbot 的护城河建立在规模 / 成本指标上,但这些指标来自公司声称,并未独立审计。
[CP024, CP017, CP004]3.5 图表
04财务
4.1 收入来源、定价模型和确认缺口
Linkerbot 唯一明确披露的收入来源,是 Linker Hand 产品家族的硬件销售;该产品线至少 13 个 SKU,覆盖 6–42 个自由度,直接卖给工业买家、人形机器人集成商和研究机构。公司也营销一层软件 / 平台能力——LinkerSkillNet 操作技能生态和 Open TeleDex 遥操作系统——但已审阅来源没有披露它们是否产生独立于硬件销售的许可、订阅或数据访问收入;本报告把它记录为开放问题,而不是默认它已经是实际收入线或只是营销包装。定价方面,独立行业跟踪方把 O6 入门型号的标价归因于约 ¥6,666(约 $920),这是已审阅来源中唯一找到的具体单价;同一来源还引用公司联合创始人预测价格可在三年内降到 ¥500(约 $70)以下——这是激进的远期成本目标,而非当前价格。高 DoF 旗舰型号(L20、L20G、L30、L30Z、O7 等)没有来源披露价格,因此目前无法拼出完整产品定价梯度;在缺少收入结构披露的情况下,也无法评估收入确认处理。[CI001, CI002, CI003, CI004, CI005, CI036]
| 收入流 | 机制 | 单位 | 当前数值 / 状态 | 质量 | 尽调问题 |
|---|---|---|---|---|---|
| 灵巧手硬件销售 | 向工业、集成商和科研买家直接销售 Linker Hand SKU | 按单位 | 截至 2025 年 12 月累计出货 10,000+ 只(公司声称) | 公司声称,部分得到第三方追踪器印证 | 索取按型号 / 细分市场拆分的销量 |
| 操作技能软件(LinkerSkillNet) | 可能是与硬件打包或并行销售的许可 / 数据访问层 | Unknown | 未披露为独立收入 | 未证实 / 待回答问题 | 确认是否存在任何独立软件收入 |
| 遥操作平台(Open TeleDex) | 可能是面向遥操作部署的平台 / 服务层 | Unknown | 未披露为独立收入 | 未证实 / 待回答问题 | 确认该系统是否带来许可或服务收入 |
| 科研 / 机构销售 | 与高校直接销售或合作(公司称包括 Stanford、MIT) | 按单位 / 按项目 | 未量化 | 公司声称 | 索取科研渠道收入占比 |
没有来源披露收入拆分;本表列举的是根据产品和渠道表述披露或可推断的收入流,并非经审计 P&L。
[CI001, CI002, CI005]| 价格 / 单位 / 合同 | 标价与实际成交价 | 折扣 / 未知项 | 来源 |
|---|---|---|---|
| Linkerbot O6:~¥6,666 (~$920) | 独立追踪器给出的标价;实际 / 批量定价未披露 | 批量折扣条款未披露 | TechBuzz China 行业追踪器 |
| 公司给出的 3 年目标:低于 ¥500 (~$70) | 前瞻性目标,不是当前标价 | 达到该价格的时间表和成本路径未披露 | TechBuzz China 引用的联合创始人说法 |
| 旗舰型号(L20、L20G、L30、L30Z、O7 等) | 未披露 | 高 DoF 层级未找到公开价格 | 未找到;linkerbot.cn/price 页面由 JS 渲染,无法提取价格表 |
只有 O6 入门层级价格得到独立来源印证;其余 13+ SKU 产品线没有披露公开价格。
[CI003, CI004, CI036]4.2 GTM 动作、销售效率和成本结构
Linkerbot 已披露的 GTM 动作,是直接销售给具名工业和研究账户(Samsung、Siemens,以及被描述为包括 Stanford 和 MIT 的研究用途),同时借助国际媒体和展会可见度,尤其是 CES 2026 首秀;已审阅来源没有披露正式经销商或分销商网络,也没有披露销售周期长度、获客成本或回本周期代理指标。成本结构方面,行业评论把 Linkerbot 的低标价部分归因于自有零部件(PEEK 工程聚合物、自研微型减速器)和制造规模,而不是披露的人工、制造费用或物流数据;没有来源披露毛利率、BOM 成本、营运资本状况,或计划资本开支,只有笼统表述称新融资用于“产能”提升。可比同业扩产阈值报道(例如 AgiBot 冲过 10,000 台)可帮助理解 Linkerbot 自己 10,000 台 / 月目标大概率需要多少资本强度,尽管没有 Linkerbot 专属资本开支数字。[CI006, CI007, CI008, CI009, CI010, CI011]
融资、制造和平台研发三个环节资本强度都高,且每个阶段的现金流可见度都低。
[CI024, CI029, CI033]4.3 公开牵引力与私有指标缺口
Linkerbot 的公开牵引力完全基于台数和产量:公司披露截至 2025 年 12 月累计出货第 10,000 只灵巧手,并正把月产量从接近 5,000 台提升到 2026 年中目标 10,000 台;另有中文报道称公司目标是在 2026 全年交付 50,000–100,000 台,相当于其 2025 年末披露累计出货量的 5–10 倍。所有这些活动都没有与财务报表级指标匹配:没有来源披露收入、ARR、GMV、活跃客户数或利用率数据。CB Insights 维护一个分析师跟踪的公司档案,说明存在一定独立金融数据提供商覆盖,但已审阅摘录没有显示具体财务字段。因此,公开记录呈现出一个结构性缺口:运营 / 台数牵引力有强且多方佐证,财务牵引力证据却完全缺席。[CI012, CI013, CI014, CI028]
| 缺失的私营公司指标 | 影响 | 精确尽调路径 |
|---|---|---|
| 收入 / ARR | 无法用任何收入倍数校准估值 | 索取管理账或经审计财务报表 |
| 毛利率 | 无法判断硬件定价策略是否可持续 | 按 SKU 索取 BOM / COGS 数据 |
| 员工人数 | 无法计算人均产出或研发强度代理指标 | 索取组织架构图 / 人员披露 |
| 账面现金和烧钱率 | 无法脱离下一轮融资评估现金跑道 | 索取资产负债表 / 现金流量表 |
| 累计融资总额 | 无法计算稀释或资本效率(每美元融资对应的收入或出货量) | 索取完整股权结构表和逐轮融资摘要 |
| 客户级收入集中度 | 无法从财务报表角度评估客户集中风险(客户 logo 层面视角见「客户」章节) | 索取头部客户收入占比 |
每一行都代表本章审阅的全部来源均未公开披露,不只是漏掉某一个来源。
[CI013, CI019, CI020, CI027, CI032]资本投入手部制造,转化为已披露的单位出货牵引,但收入 / 利润率结果仍未披露。
节点标签反映已披露活动;收入 / 毛利节点明确标注为未披露,而非估算。
[CI001, CI002, CI006, CI012]4.4 资本充足性和融资依赖
在公司概览章节已记录的历史逐轮融资脉络基础上(本章不重复),本章与资本充足性直接相关的证据显示:Linkerbot 在 2026 年 2 月 12 日前后完成约 RMB 1.5B(约 $210M)的 Series B,由 Daode Investment 和 Shengshi Investment 领投,另有很长的参投方名单,披露募资用途聚焦核心研发、产能扩张和全栈操作平台能力。到 2026 年 5 月上旬,公司又以约 $3B 估值完成 ‘Series B+’,投资方包括 Ant Group、HongShan Group、Zhongguancun Science Park Fund、Bank of China Asset Management 和 Fosun Capital,并披露下一轮目标约 $6B。另一个单列证据是,上市公司 Dongfang Precision(002611.SZ)披露直接持有 Linkerbot 约 4.36% 股权,并在 2026 年 6 月投资者关系回复中引用其 2025 年年报确认持股未变;这是一条罕见的、达到财务披露级别的 Linkerbot 股权结构证据,因为它来自上市公司自身监管披露义务,而不是私营公司新闻稿。没有来源对 2026 年 2 月 Series B 轮金额与 2026 年 5 月 Series B+ 估值隐含的资本规模作对账,因此两次事件之间的稀释和资本效率无法完全拆解。没有来源披露账上现金、月度现金消耗、资金可支撑时间、具体下一轮触发里程碑,或任何债务 / 项目融资义务;中国国家发展和改革委员会 2026 年关于国内 150 多家人形机器人公司行业性产能过剩的警示,是解读 Linkerbot 快速、反复融资节奏时必须纳入的负面宏观信号。[CI015, CI016, CI017, CI018, CI019, CI020]
| 指标 | 数值 / 空值 | 置信度 | 为什么重要 | 尽调问题 |
|---|---|---|---|---|
| 单件毛利率 | 低 | 决定当前标价是否可持续,还是增长阶段 / 亏损引流定价 | 索取按 SKU 拆分的 COGS 和 BOM 成本 | |
| 获客成本 | 低 | 评估商业化效率和回本周期所需 | 索取销售 / 营销支出和新账户数据 | |
| 质保 / 缺陷 / 退货率 | 低 | 影响真实到岸销货成本 | 索取服务与质保成本数据 | |
| 人均收入(代理指标) | 低 | 未披露员工人数,因此无法计算该代理指标 | 索取员工人数和收入披露 | |
| 合同期限 / 复购率 | 低 | 影响收入耐久性假设 | 索取客户留存 / 复购数据 |
已审阅公开来源均未披露任何单位经济指标;这里的 null 代表真实信息缺口,不是零值。
[CI008, CI025, CI026, CI027]| 项目 | 数值 / 状态 | 来源 | 备注 |
|---|---|---|---|
| 账面现金 | 未披露 | 未披露资产负债表或现金头寸 | |
| 月度烧钱 | 未披露 | 未找到运营成本披露 | |
| 现金跑道(月) | 未披露 | 缺少现金和烧钱数据,无法计算 | |
| 2026 年 2 月 B 轮 | ~RMB 1.5B (~$210M);由 Daode Investment 和 Shengshi Investment 领投 | 163.com | 用途:核心研发、产能、全栈基础能力 |
| 2026 年 5 月「B+ 轮」估值 | ~$3B;投资方包括 Ant Group、HongShan、Zhongguancun Science Park Fund、BOC Asset Management、Fosun Capital | BusinessWorld / Straits Times / Yahoo Finance(Reuters 通稿) | 尚未与 2026 年 2 月 B 轮融资规模完全对齐 |
| 下一轮目标估值 | ~$6B(时间 / 工具未披露) | BusinessWorld / Straits Times | 「公司概览」年表保留完整的历史逐轮融资时间线;本表只重述与资本充足性有关的事实 |
| 上市公司持股 | Dongfang Precision (002611.SZ) 持有约 4.36% 直接股权,自 FY2025 年报以来未变 | Tencent News / East Money(投资者关系披露) | 公开市场披露证实至少一名机构 / 战略股东仍在持股 |
| 债务 / 项目融资义务 | 未披露 | 没有来源显示存在任何非股权融资 |
本表只重述与资本充足性有关、且带有本地财务 sourceRefs 的融资事实;完整的历史逐轮融资记录在「公司概览」里程碑表中。
[CI015, CI016, CI017, CI018, CI019, CI020]真正的单位经济性桥所需的每个量化输入都未披露;节点只展示目标结构。
由于 COGS、毛利率和 opex 均未披露,本图记录的是目标桥接结构,而非填入数值;仅作说明。
[CI003, CI008, CI025, CI026]已披露轮次规模与 Series B+ 估值隐含的累计募资额之间差距很大,且尚未勾稽。
Series B 行使用约 7.1-7.9 CNY/USD 区间,将已披露的 ~RMB 1.5B 折算为 USD;Series B+ 隐含资本行是说明性、未勾稽的占位区间,不是已披露或经审计数字——见资本充足性证据缺口。
[CI015, CI018, CI030, CI031]4.5 财务结论:收入质量、利润率路径和尽调阻断项
合在一起看,Linkerbot 的公开记录同时包含强且多方佐证的台数出货牵引力和融资轮活动,也几乎完全缺少收入、利润率、成本结构和资产负债表披露;这符合快速扩张的私营硬件公司常见模式,但会实质性阻断任何独立财务估值。收入质量无法评估,因为没有来源按收入流、客群或地域拆分收入。利润率路径也不清楚,因为毛利率、BOM 成本、保修 / 缺陷成本全部未披露;联合创始人所称未来数年走向 $70 以下单价,既可能反映可信的制造成本曲线,也可能是由持续股权融资支持的有意亏损扩张策略——公开记录无法区分两者。可比同业制造扩产报道显示资本强度很高;Linkerbot 自身融资依赖(三次或更多股权融资发生在不到 18 个月内)大概率会持续,直到硬件利润率和 / 或未披露的软件 / 数据收入层成熟到足以内部供血,除非这一点永远无法实现。近期最具体的尽调阻断项,是没有任何经审计财务报表(可比同业 Unitree 已通过自身 IPO 招股书跨过这道门槛)、2026 年 2 月 Series B 轮金额与 2026 年 5 月 Series B+ 估值之间缺少对账,以及员工人数、现金状况和现金消耗率披露完全缺席。[CI029, CI032, CI033, CI034, CI035]
4.6 图表
05产品与技术
5.1 产品定义、工作流和模块图谱
从客户工作流看,Linkerbot 的产品是灵巧机器人手末端执行器:集成商把它安装到机器人臂或人形机器人平台上,再通过 ROS 或 ROS2 驱动栈和 Python/C++ SDK 控制;它不是客户直接购买的完整机器人。已记录的集成流程——在 Ubuntu 22.04、ROS Humble 和 Python 3.10+ 环境安装 SDK,通过 setting.yaml 文件配置手部参数,再用 ROS topics 或 LinkerHandApi Python 类驱动手部——可以在 Linkerbot 自己托管于 GitHub 的 SDK 仓库中直接看到。公司官方站点地图至少列出 13 个不同产品页面(L6、L10、L20、L20G、L30、L30Z、O6、O7、D、J、R、W、YCB),但 ROS2 SDK 自己的 README 记录的硬件支持清单不同且只部分重叠(O6、L6、L7、O7、L10、L20、G20、L21);本报告把这视为命名和覆盖范围不一致,而不是将其默认抹平。公司材料披露的旗舰规格包括:L30 面向精密装配工作约 ±0.2mm 重复精度;O6 重 370g、抓握载荷 50kg,归因于 PEEK 工程聚合物和自研微型减速器。[CE001, CE002, CE003, CE004, CE005, CE006]
| 模块 / 产品线 | 用户 | 状态 / 成熟度 | 差异化 | 尽调缺口 |
|---|---|---|---|---|
| O6(6 DoF,370g,50kg 负载) | 对成本敏感的工业买家 | 已出货 / 量产(公司称) | PEEK 聚合物 + 微型减速器,带来高载荷重量比 | 未审阅独立载荷测试数据 |
| L20(工业级,Super Electric Cylinder) | 工业重载自动化 | 已出货 / 量产(公司称) | 90% 驱动效率、200N 推力、100 万次以上循环测试寿命(公司称) | 未审阅独立耐久性测试 |
| L30(精密级,±0.2mm 重复定位精度) | 精密装配买家 | 已出货 / 量产(公司称) | 面向精细装配任务的高精度定位 | 未审阅独立精度基准 |
| Linker Craftsman 工作站 | 需要完整操作工作站的集成商 | CES 2026 亮相(2026 年 1 月) | 打包手部 + 工作站,而不只是单个手部 | 除亮相外,未确认出货 / 量产状态 |
| SDK 支持子集(GitHub README 中的 O6、L6、L7、O7、L10、L20、G20、L21) | 开发者 / 集成商 | 开源 SDK 持续维护(ROS、ROS2、Python、C++) | 在已审阅的中国灵巧手厂商中,开放开发者工具最完整 | 型号命名和覆盖范围未完全匹配官网站点地图的 SKU 清单 |
| 其余站点地图 SKU(L20G、L30Z、D、J、R、W、YCB) | 多类用户(按产品页命名) | 官网列出;页面内容由 JavaScript 渲染,可提取细节很少 | 扩大产品线宽度叙事 | 已审阅来源未确认这些型号的规格和 SDK 覆盖 |
产品 / 模块状态来自公司说法和已审阅的 GitHub SDK 文档,不是独立第三方测试数据。
[CE003, CE004, CE005, CE006, CE008, CE014]5.2 架构和运行模式
Linkerbot 的手部在一个产品家族内覆盖三种驱动架构——直驱、连杆和腱驱动;独立行业跟踪方指出,相较多数单架构对手,这种技术路线覆盖异常宽。公司材料称工业级 L20 使用自有 ‘Super Electric Cylinder’,可提供 90% 驱动效率、200N 丝杆端推力,并经过超过 100 万次循环寿命测试;多模态 “sensing skin” 触觉层也在整个手部家族中营销。硬件层之上,是托管在 linker-bot 组织 GitHub 下的开源 ROS/ROS2 SDK、URDF 模型和模拟器(linker-sim)——这些是真正可观察的开发者工具,不是营销说法。相比之下,没有按名称找到 ‘Open TeleDex’ 或 ‘LinkerSkillNet’ 的 GitHub 仓库或技术文档;它们是公司最常营销的两个高层平台名,最接近的仓库只是一个通用遥操作臂工具。这说明这些产品名可能覆盖更窄命名的底层代码,也可能尚未开源。公司没有发布一张图,把硬件、固件、开源 SDK 层和这两个专有平台层连起来;因此,本章架构只能从分散披露中拼出。[CE007, CE008, CE009, CE010, CE027, CE028]
| 层级 / 组件 | 作用 | 依赖 | 风险 |
|---|---|---|---|
| 驱动(直驱 / 连杆 / 腱绳驱动) | 覆盖产品家族的物理运动生成 | 自有机械设计 + 零部件供应(电机、齿轮箱) | 未披露供应商冗余;是否存在单一供应商依赖也未确认 |
| 感知皮肤(多模态触觉感知) | 为操作提供力 / 触觉反馈 | 内部传感器设计(按公司说法) | 未审阅独立灵敏度 / 耐久性数据 |
| ROS / ROS2 驱动 + LinkerHandApi SDK | 面向集成商的软件控制接口 | Ubuntu 22.04、ROS Humble、Python 3.10+(据 GitHub README) | 集成周期被绑定到较新的特定 OS / ROS 版本组合 |
| URDF + 仿真器(linker-sim) | 仿真和运动规划支持 | 标准机器人工具链(如 Gazebo / MoveIt 兼容格式) | 已审阅来源未独立验证 sim-to-real 迁移保真度 |
| Open TeleDex 遥操作层 | 人在环数据采集和远程控制 | 未确认是开源仓库;很可能为自有系统 | 这一层缺乏独立技术可见度 |
| LinkerSkillNet 操作技能软件 | 更高层的操作技能库 / 数据集 | 自有系统,建立在采集到的遥操作 / 操作数据之上 | 未披露数据权利、隐私或模型训练使用政策 |
架构由 GitHub、新闻稿和产品页披露拼出;未找到公司发布的单一架构图。
[CE007, CE009, CE010, CE011, CE012, CE027]Linkerbot 技术栈覆盖硬件驱动 / 传感、设备驱动 / SDK 层,以及披露深度不一的上层平台软件。
层边界来自不同披露(GitHub 仓库、新闻稿、产品页)推断;公司未发布单一架构图。
[CE007, CE009, CE011, CE027, CE028]5.3 部署、集成、可靠性和路线图
部署和集成方面,Linkerbot 提供了真实、可检查的工具:多个 SDK 语言 / 框架变体(Python、C++、ROS1、ROS2)、兼容标准仿真和运动规划工具的 URDF 模型仓库,以及专用模拟器;截至报告日,这些都支持 Ubuntu 22.04 和 ROS Humble。已披露的具名客户 / 用户覆盖工业生产(公司表述中的 Samsung、Siemens)和研究机构(Stanford、MIT),说明部署不止内部演示;但已审阅来源没有记录任何具体产线集成案例及可测量的前后对比结果,也没有任何量产型号的现场可靠性数据(MTBF、缺陷率、退货率)。除公司 JavaScript 渲染网站上的一般 FAQ 页面外,没有找到正式客户支持 SLA 或保修条款。路线图方面,Linkerbot 通过展会节点释放里程碑,而不是发布公开版本日历:CES 2026 首秀推出完整 Linker Hand 产品线和覆盖 11–42 DoF 的 ‘Linker Craftsman’ 工作站;除此之外,没有找到前瞻路线图(具名下一代型号、目标日期)。[CE011, CE012, CE013, CE014, CE015, CE023]
| 日期 / 阶段 | 功能 / 里程碑 | 状态 | 含义 | 来源 |
|---|---|---|---|---|
| 2025-12 | 累计第 10,000 只手部出货;披露自 2025 年 4 月以来融资 $150M | 已出货 | 证实公司已进入量产阶段,而不只是原型 | PR Newswire |
| 2026-01 (CES 2026) | 完整 Linker Hand 产品线 +「Linker Craftsman」工作站亮相,覆盖 11-42 DoF | 亮相 / 展示 | 将可服务用例从单个手部扩大到更广范围 | PR Newswire / RoboticsTomorrow |
| 持续进行(GitHub,未注明日期) | ROS、ROS2、Python、C++ SDK,加上 URDF 模型和仿真器仍在维护 | 文档持续维护 | 释放出持续投入开发者工具的信号 | GitHub(linker-bot org) |
| 未披露 | 具名下一代型号或发布日期路线图 | 未找到 | 未找到 CES 2026 亮相之后的前瞻路线图 | 未找到 |
路线图可见度仅限于已宣布里程碑;未在任何已审阅来源中找到前瞻发布日历。
[CE014, CE015, CE028]典型集成商从硬件采购走到 SDK 集成,再到生产部署;初始集成之后,支持 / 可靠性可见度下降。
[CE001, CE002, CE011, CE013, CE023]Linkerbot 技术栈依赖 OS / ROS 版本组合、未披露的组件供应商,以及自有专有遥操作 / 技能层。
已审阅来源均未披露组件供应商身份;节点代表类别,不是具名第三方。
[CE011, CE012, CE018, CE028]5.4 差异化、信任和合规
Linkerbot 对外讲的技术差异化,主要落在三点:一个产品家族覆盖三类主流驱动方案(多数对手通常只押一种);材料和制造自有化(PEEK 聚合物、自研微型减速器);以及靠成本 / 规模领先而非单项可专利突破拿到的 80%+ 市占率。已审阅来源都没有披露 Linkerbot 方法对应的专利组合、申请或已授权 IP 保护,也没有来源披露制造基地数量或零部件供应商细节。信任与合规披露更薄:没有来源显示 Linkerbot 自身持有正式安全或质量认证(如 ISO 10218 或 ISO/TS 15066),尽管公司的 ROS SDK README 为硬件操作写了面向开发者的非正式安全提示。没有找到针对通过 LinkerSkillNet 或 Open TeleDex 采集的操作技能或遥操作数据的数据隐私 / 安全政策,也没有披露制造质量控制流程(来料 / 制程 / 出货 QA)。中国国家发改委 2026 年对 150 多家国内人形机器人公司产品过度重复的警示,是判断全行业快速 SKU 扩张——包括 Linkerbot 的 13+ 个 SKU——是否配套了相应质量和安全投入的负面信号。整体看,开源开发者工具确实活跃,但认证、IP、可靠性数据几乎完全缺位,说明 Linkerbot 目前的公开披露策略更偏向开发者采用和媒体曝光,而不是正式企业信任文档。[CE016, CE017, CE018, CE019, CE020, CE021]
| 用户任务 | 当前工作流(不用 Linkerbot) | Linkerbot 方案 | 可衡量收益(按公司说法) | 限制 |
|---|---|---|---|---|
| 精密电子 / 工业装配 | 人工装配或低自由度夹爪自动化 | L30 手(±0.2mm 重复定位精度)通过 ROS2 SDK 集成 | 公司称精度适合精细装配 | 未审阅独立吞吐量或缺陷率基准 |
| 工业重载搬运 | 刚性夹爪或人工搬运 | L20(200N 推力、100 万次以上循环寿命)通过 ROS2 SDK 接入 | 公司称耐久性和力冗余充足 | 未审阅独立载荷循环测试 |
| 研究型操作实验 | 高成本研究级手部(如 Shadow Robot、Allegro Hand) | O6 / 轻量型号标价约 ~$920(据行业追踪器) | 以更低成本获得高自由度研究硬件 | 未审阅专门验证 Linkerbot 手部的同行评议研究 |
| 面向具身 AI 训练的遥操作数据采集 | 自建遥操作装置 | Open TeleDex 遥操作系统 | 公司称数据采集工作流更顺 | 未找到具名仓库或技术文档来确认 Open TeleDex 的确切架构 |
除非另有说明,收益均来自公司说法;未找到任何所列用例的独立基准研究。
[CE001, CE002, CE005, CE008, CE009, CE010]| 控制 / 认证 / 质量指标 | 状态 | 范围 | 缺口 |
|---|---|---|---|
| ISO 10218 / ISO-TS 15066 工业 / 协作机器人安全认证 | 未披露 Linkerbot 已持有 | 行业层面存在标准;公司特定认证未确认 | 直接向 Linkerbot 索取认证文件 |
| ROS SDK 安全提示 | GitHub README 中存在(非正式、面向开发者) | 仅为软件 / 操作指引,不是第三方认证 | 确认这些提示背后是否有正式安全测试 |
| 遥操作和技能数据的数据隐私 / 安全政策 | 未披露 | 关系到 LinkerSkillNet 和 Open TeleDex 数据采集 | 直接向公司索取数据处理和隐私政策 |
| 制造质量控制流程(来料 / 过程 / 出货 QA) | 未披露 | 适用于所有已出货 SKU | 尽调时索取质量管理文件(如 ISO 9001 状态) |
| 行业层面的产能过剩 / 质量风险警告(NDRC,2026) | 在行业层面公开披露,不针对单一公司 | 广泛适用于 150 多家中国人形机器人公司 | 跟踪是否出现针对公司的监管发现 |
未在任何已审阅来源中找到专门针对 Linkerbot 的正式第三方认证;本表明确记录缺失,而不是假定其合规。
[CE019, CE020, CE021, CE022, CE032]硬件和开源开发者工具是披露最充分、成熟度最高的层;认证、IP 和可靠性数据最不成熟,披露也最少。
[CE016, CE017, CE019, CE035]5.5 要点证据
06客户
6.1 客户基础分层
Linkerbot 已披露客户基础至少可分为三类。工业和电子制造客户——多份来源点名 Samsung Electronics 和 Siemens——属于买方、用户、付款方通常都落在同一企业采购 / 自动化职能的群体。研究机构——公司新闻材料中的 Stanford、MIT、Cambridge,以及独立媒体报道中的 University of Hong Kong——资金来自科研经费和实验室预算,而不是直接商业采购;其价值更多在声誉和管线,而非直接收入。人形机器人集成商构成隐含的第三类买家(因为 Linkerbot 的手部产品设计用于安装在机器人平台上),但已审阅来源没有点名任何集成商是 Linkerbot 的直接客户。除上述账户外,没有来源披露地理分布、公司规模 / 收入段分层,或按细分市场拆出的完整客户数量。因此,分层分析只能锚定少数具名 logo,而不是完整账户地图。[CU001, CU002, CU003, CU004]
| 分群 | 买方 / 用户 / 付款方 | 用例 | 规模 | 收入 / 战略价值 | 缺口 |
|---|---|---|---|---|---|
| 工业 / 电子制造 | 企业采购 = 买方 / 付款方;产线工程师 = 用户 | 精密装配、产线自动化 | 未披露 | 具名客户:Samsung Electronics、Siemens | 未披露客户级收入或销量 |
| 研究机构 | 大学实验室(Stanford、MIT、Cambridge、University of Hong Kong)= 用户;科研经费 = 付款方 | 操作研究、数据集采集、演示 | 未披露 | 声誉 / 管线价值;不是主要收入分群 | 未披露研究渠道收入占比 |
| 人形机器人集成商(隐含) | OEM 工程团队 = 买方 / 用户;研发 / BOM 预算 = 付款方 | 为完整人形机器人采购部件 | 未披露 | 已审阅来源未直接将其列为客户 | 未找到已确认、具名的集成商客户 |
分群由公司和新闻披露中点名的客户拼出;没有来源按分群给出账户数量或收入结构拆分。
[CU001, CU002, CU003, CU004]6.2 采用轨迹
Linkerbot 的采用轨迹几乎完全靠出货量和产能规模披露来支撑,而不是账户级指标。月订单量据报到 2025 年 11 月已超过 1,000 台,累计出货到 2025 年 12 月超过 10,000 台;到 2026 年 5 月,公司称月产能正从接近 5,000 台爬向 10,000 台目标。另一篇中文报道称,公司目标是在 2026 年全年交付 50,000 至 100,000 台——这一目标显著高于已披露的 2025 年底累计水平,但没有来源确认年中交付节奏是否跟上该目标。出货量之外,Linkerbot 还在第九届中关村国际前沿科技大赛中夺冠(2026 年 3 月 26 日公布),参赛项目超过 2,100 个,属于第三方技术验证,不同于直接销售或使用指标。已审阅来源没有把这些出货活动换算成独立客户账户数、部署地点数或利用率,所以采用轨迹仍是产能规模故事,而不是客户账户故事。[CU005, CU006, CU007, CU008]
| 指标 | 数值 | 日期 | 来源 | 置信度 | 含义 | 缺失分母 |
|---|---|---|---|---|---|---|
| 月订单量 | >1,000 units/month | 2025-11 | China Biz Insider | 中 | 首次披露的订单量门槛 | 缺少不同客户数量分母 |
| 累计手部出货量 | 10,000+ | 2025-12 | PR Newswire | 高 | 证实公司已进入量产阶段 | 缺少复购客户与新客户拆分 |
| 月产量(当前) | ~5,000 units/month | 2026-05 | BusinessWorld / Straits Times | 高 | 冲刺月产 10,000 单位目标前的扩产基线 | 缺少产能利用率或积压订单数据 |
| 月产量(目标) | 10,000 units/month | 2026-05 | BusinessWorld / Straits Times | 高 | 显示继续扩产的雄心 | 缺少达到目标的确认日期 |
| 2026 全年交付目标 | 50,000-100,000 台 | 2026-02 | 163.com | 中 | 隐含较 2025 年末累计出货量增长 5-10 倍 | 未找到年中产量节奏确认 |
所有数字都来自公司说法或第三方报道;未找到独立审计的出货或账户数据。
[CU005, CU006, CU007, CU008]已披露证据集中在认知和发货阶段;扩张与留存阶段没有披露数据。
漏斗数值只是示意比例,反映各阶段已披露证据的相对深度,不是实测转化率。
[CU009, CU014, CU018, CU028]6.3 具名客户证据:生产部署还是试点、证据是否新鲜
在已审阅来源中找到六个具名客户 / 机构:工业侧的 Samsung Electronics 和 Siemens,研究侧的 Stanford、MIT、Cambridge、University of Hong Kong。Samsung Electronics 证据最强,在三个相对独立的来源中反复出现(China Biz Insider、PR Newswire、Interesting Engineering);Siemens 和 Stanford 各有两个来源交叉印证;MIT 和 Cambridge 只出现在 Linkerbot 自己的新闻稿中,未在本次审阅来源里获得独立确认;University of Hong Kong 只出现在一个独立媒体来源中,且未被 Linkerbot 自己已审阅材料提及。关键是,六个客户 / 机构没有一个被来源说明关系性质是生产部署、试点,还是一次性演示 / 设备借用;也没有来源给出任何账户层面的量化结果(周期缩短、缺陷率变化、成本节省)——本次找到的全部客户证据都是 logo 级,而非结果级。这些证据时间都在 2025 年 11 月至 2026 年 5 月之间,满足 2026 年中报告的新鲜度要求,但没有来源明确重新确认这些关系在报告日仍然活跃。[CU009, CU010, CU011, CU012, CU013, CU014]
| 客户 | 分群 | 部署 / 用例 | 量产还是试点 | 结果 | 限制 |
|---|---|---|---|---|---|
| Samsung Electronics | 工业 / 电子制造 | 精密装配 / 自动化(隐含) | 未说明 | 未量化 | 多个来源点名,但没有结果指标,也未确认量产 / 试点 |
| Siemens | 工业 | 自动化 / 制造(隐含) | 未说明 | 未量化 | 公司和独立媒体均点名,但无部署细节 |
| Stanford University | 研究 | 操作研究(隐含) | 研究 / 演示用途,不是商业量产 | 未量化 | 未确认具名实验室、论文或具体项目 |
| MIT | 研究 | 操作研究(隐含) | 研究 / 演示用途,不是商业量产 | 未量化 | 仅为公司说法;已审阅来源未找到独立佐证 |
| Cambridge | 研究 | 操控研究(隐含) | 研究 / 演示用途,不是商业化生产 | 未量化 | 仅为公司说法;已审阅来源未找到独立佐证 |
| University of Hong Kong(研究机构) | 研究 | 操控研究(隐含) | 研究 / 演示用途,不是商业化生产 | 未量化 | 仅由一家独立媒体来源点名,公司已审阅新闻稿未点名 |
所有具名客户证据都停留在 logo 级别(点名账户),不是结果级别(可衡量成果);本表记录的是可获得证据层级, 不是案例研究结论。
[CU009, CU010, CU011, CU012, CU013, CU014]工业和科研客户的采用入口不同,但最终都靠直销或媒体报道发现产品,而不是通过渠道伙伴导入。
[CU001, CU002, CU021, CU026]所有具名客户中,证据质量在客户标识确认上最强,在结果具体性和生产状态确认上最弱。
[CU009, CU010, CU011, CU012, CU034]6.4 留存、韧性、扩张与集中度风险
留存和关系韧性方面,公开记录为空:已审阅来源没有披露 Linkerbot 的净收入留存、总收入留存、流失率、续约率、合同期限或任何客户满意度指标,也没有来源记录任何具名账户的复购行为。因此,订单量上升不能归因于新客增长、复购 / 扩张订单,或两者混合。扩张和集中度方面,没有来源披露最大客户或前五大客户收入占比,也没有披露独立客户 / logo 总数,导致集中度风险无法按总账户基础占比评估,只能看绝对的具名客户。基于 Linkerbot 披露的直销与展会获客动作,渠道 / 伙伴依赖似乎有限;不过两个国际 B2B 贸易目录条目——HKTDC 供应商店铺和 Hannover Messe 参展商页面——截至报告日均返回 not-found,对可能通过这些渠道发现 Linkerbot 的国际买家构成轻微采购摩擦信号。行业层面,中国国家发改委 2026 年对 150 多家国内人形机器人公司的警示是负面信号:竞争场越来越拥挤,可能带来客户切换压力,而 Linkerbot 当前稀薄的留存和集中度披露无法评估这种压力。整体看,客户证据画像是一边有可信、由多来源交叉印证的具名客户 logo,另一边几乎没有留存、集中度、满意度和结果级证据——更像一家早期爬坡的硬件公司,而不是成熟企业供应商。[CU016, CU017, CU018, CU019, CU020, CU021]
| 指标 | 数值 / null | 分群 | 置信度 | 尽调请求 |
|---|---|---|---|---|
| 净收入留存(NRR) | 全部 | 低 | 索取分 cohort 收入留存数据 | |
| 毛收入留存(GRR) | 全部 | 低 | 索取分 cohort 收入留存数据 | |
| 流失率 | 全部 | 低 | 索取客户流失追踪数据 | |
| 续约 / 复购率 | 全部 | 低 | 索取按账户划分的订单历史数据 | |
| 合同期限 | 工业 | 低 | 索取标准合同 / 条款文件 | |
| 客户满意度(NPS 或调研) | 全部 | 低 | 索取内部或第三方满意度调研数据 |
已审阅公开来源均未披露任何留存 / 满意度字段;null 代表真实信息缺口,不是零值。
[CU016, CU017, CU018, CU019]| 扩张驱动因素 | 集中度风险 | 影响 | 尽调路径 |
|---|---|---|---|
| 月订单量上升(>1,000,目标 10,000/月) | 未披露最大客户或前 5 大客户收入占比 | 无法判断增长是广泛分散,还是集中在少数大客户 | 索取客户级收入或销量拆分 |
| 具名多分群客户名单(工业 + 研究) | 未披露去重客户 / logo 总数 | 无法按总账户数占比计算集中度 | 索取活跃客户总数 |
| 直销 + 展会曝光(CES 2026) | 两个国际 B2B 贸易目录条目(HKTDC、Hannover Messe)返回未找到结果 | 使用这些渠道的潜在国际买家会遇到轻微采购摩擦信号 | 确认当前可用的国际采购 / 贸易目录条目 |
| 全行业产能过剩警告(NDRC,2026) | 行业广泛整合可能挤压所有竞争者的客户关系 | 如果竞争对手在价格或可靠性上压价,可能间接影响 Linkerbot 的客户留存 | 跟踪政策和竞争对手整合信号 |
鉴于缺少任何账户级收入或销量披露,集中度风险只能定性评估。
[CU020, CU021, CU022, CU023, CU028, CU032]Samsung Electronics 的独立来源印证最深;University of Hong Kong 与 MIT/Cambridge 各自只靠一个来源支撑。
数值统计已审阅来源中点名各客户的不同来源数量(公司或独立来源),不是销售额、收入或满意度指标。
[CU009, CU010, CU011, CU012, CU013, CU015]6.5 要点证据
07风险
7.1 按严重度排序的风险概览
从监管、运营、伙伴和人员维度看,Linkerbot 最高严重度风险包括:在没有披露资本开支计划的情况下,可能无法兑现 10,000 台 / 月(以及另有报道的 2026 全年 50,000-100,000 台)产能目标;美国联邦采购禁令(GUARD Act 和 American Security Robotics Act)可能切断美国链路上的中国制造机器人组件需求;关键人和接班风险集中在唯一公开具名高管身上;资本提供方集中在少数大型、部分带国资背景的机构投资者手中。每项风险都有合理投资影响:制造爬坡不及预期会压低近期收入增长;监管会缩小美国可服务市场;人员风险可能造成领导层突然断档;资本集中则会在投资者情绪转向时带来融资脆弱性。缓释成熟度普遍低:已审阅来源没有记录 Linkerbot 针对任何顶级风险披露具体应急方案、保险政策或风险管理安排;对一家估值数十亿美元的公司来说,这本身就是显著缺口。[CR001, CR016, CR020, CR025, CR034, CR041]
7.2 监管与法律风险
Linkerbot 同时面对美国和中国两侧的监管风险。美国方面,议员在 2026 年 3 月提出两党共同支持的 GUARD Act 和 American Security Robotics Act,目标是禁止联邦政府采购和使用中国制造的人形及地面机器人;没有已审阅来源确认范围是否延伸到嵌入非中国品牌机器人中的灵巧手等组件,因此这一风险对 Linkerbot 的实际触达仍不确定,但合理存在。另行来看,BIS 2026 年逐案审查 AI 芯片许可证的流程,以及 2025 年 1 月更广泛的先进计算物项和 AI 模型权重出口管制,都给任何采购或出口先进算力的具身 AI 硬件 / 软件厂商增加合规义务。一家律所分析称,人形机器人整体正在既有产品责任和安全法律下进入「风险新时代」,但已审阅来源没有记录任何针对 Linkerbot 的诉讼、执法行动或 IP 争议。中国方面,全国人形机器人标准化技术委员会正在组建,Unitree 和 AgiBot 高管为成员,但没有已审阅来源确认 Linkerbot 进入委员会;这带来一个风险:有约束力的技术标准可能在缺少其直接参与的情况下成形。与此同时,中国国家发改委在 2026 年公开警示 150 多家国内人形机器人公司存在产能过剩和投机泡沫,是未来政策可能收紧的行业信号。没有来源确认 Linkerbot 是否持有 ISO 10218 或 ISO/TS 15066 安全认证;EU AI Act 也给任何欧洲销售增加了另一层目前尚未确认的合规要求。[CR001, CR002, CR003, CR004, CR005, CR006]
| 规则 / 许可 / 案件 | 辖区 | 状态 | 可能性 | 严重性 | 缓释措施 | 剩余暴露 | 尽调路径 |
|---|---|---|---|---|---|---|---|
| GUARD Act / American Security Robotics Act(针对中国机器人的联邦采购禁令) | 美国 | 2026 年 3 月提出;尚未立法生效 | 中 | 高 | Linkerbot 未披露 | 若以零部件层面范围生效,暴露为高 | 跟踪法案文本和最终生效范围 |
| BIS AI 芯片许可审查流程 | 美国(对中国相关供应链有域外影响) | 已生效,2026 年逐案审查 | 高(持续) | 中 | Linkerbot 未披露 | 中 — 影响先进计算部件,不影响核心手部机械结构 | 确认 Linkerbot 在计算 / 芯片采购上的暴露 |
| 先进计算物项和 AI 模型权重出口管制 | 美国 | 自 2025 年起生效 | 高(持续) | 中 | Linkerbot 未披露 | 中 — 关系到任何 AI 模型出口或训练数据流动 | 确认 LinkerSkillNet 数据 / 模型是否跨越美国管制边界 |
| 中国人形机器人国家标准委员会 | 中国 | 2026 年处于起草阶段;未确认 Linkerbot 是委员会成员 | 中 | 中 | Linkerbot 未披露 | 中 — 标准可能在没有 Linkerbot 直接参与的情况下成形 | 确认 Linkerbot 是否有委员会席位或意见反馈流程入口 |
| ISO 10218 / ISO-TS 15066 安全认证状态 | 全球 / 全行业 | Linkerbot 认证状态未确认 | 中 | 中 | 仅有非正式 ROS SDK 安全说明(见产品与技术章节) | 中 — 可能挡住要求认证的大型工业买家 | 直接向 Linkerbot 索取认证文件 |
| NDRC 产能过剩 / 泡沫警告 | 中国 | 2026 年发布公开警告;尚无具体执法行动 | 中 | 中 | Linkerbot 未披露 | 中 — 可能预示产业政策或补贴规则收紧 | 监测 NDRC 是否有超出表态的具体政策动作 |
各行按严重性排序(最高在前);多数行反映全行业监管风险,而非已确认的 Linkerbot 专项执法行动。
[CR001, CR002, CR003, CR004, CR005, CR006]7.3 运营、质量与伙伴 / 依赖风险
运营风险上,最实质缺口在于 Linkerbot 披露的产能野心——从约 5,000 台 / 月爬坡到目标 10,000 台 / 月,另有报道称 2026 全年交付目标为 50,000-100,000 台——与完全缺失的制造产能或资本开支计划之间的落差。没有任何产品披露现场可靠性数据(缺陷率、MTBF)、召回历史或网络安全实践;公司的软件栈还绑定特定 OS / 框架(Ubuntu 22.04、ROS Humble),如果这一组合落后于更新工具,就会变成运营风险。一个行业跟踪方认为,随着独立手部产品进入者(AgiBot、Zhaowei)用价格竞争,该品类正走向商品化,这会压缩 Linkerbot 以成本 / 规模为基础的护城河;同时,Unitree 与 AgiBot 围绕出货领先说法的公开争议,也暴露了可信度风险,同样适用于 Linkerbot 自己未经审计的 80%+ 份额说法。伙伴和依赖风险上,Linkerbot 的资本基础集中在少数大型、部分带国资背景的机构投资者(Ant Group、HongShan Group、Zhongguancun Science Park Fund、Bank of China Asset Management、Fosun Capital),另有一个披露的公开市场战略持有人(Dongfang Precision,约 4.36%);这一小圈子里任何投资者情绪变化,都可能实质影响未来融资。客户集中度无法量化,因为总客户数未披露,公开识别的具名账户也只有少数。两个国际贸易目录条目(HKTDC、Hannover Messe)目前找不到,是轻微渠道依赖缺口;更广泛的半导体供应链则暴露在升级中的美国出口管制政策下。[CR013, CR014, CR015, CR016, CR017, CR018]
| 失效模式 | 可能性 | 严重性 | 缓释成熟度 | 剩余暴露 | 未解决缺口 |
|---|---|---|---|---|---|
| 制造爬坡不足(5,000 -> 10,000+ 台/月,或 2026 全年 50,000-100,000 台目标) | 中 | 高 | 未披露 | 高 — 未披露资本开支或扩产计划 | 索取制造产能和资本开支计划 |
| 现场可靠性失效(缺陷率、MTBF)未披露 | unknown | 高 | 仅有非正式 ROS SDK 安全说明 | 高 — 无法排除未披露的可靠性问题 | 索取现场可靠性 / 质保索赔数据 |
| SDK、固件或遥操作数据管线存在网络安全漏洞 | unknown | 中 | 未披露 | 中 — 未发现渗透测试或漏洞披露政策 | 索取安全审计文件 |
| 独立灵巧手新进入者(AgiBot、Zhaowei)带来商品化 / 利润率压缩 | 中 | 中 | 公司声称成本 / 规模领先,但未经审计 | 中 — 可能随时间压缩 Linkerbot 定价权 | 按季度跟踪竞争对手价格动作 |
| 未经审计的自报市场份额 / 出货量带来可信度风险 | 中 | 中 | 未披露 | 中 — 同业已围绕类似说法出现公开争议 | 委托独立市场份额审计 |
各行按严重性排序(最高在前);在没有来源披露足够信息来估算概率时,可能性标为「未知」。
[CR013, CR014, CR015, CR016, CR017, CR018]| 依赖项 | 交易对手 | 角色 | 集中度 | 失效场景 | 严重性 | 缓释措施 | 剩余暴露 |
|---|---|---|---|---|---|---|---|
| 机构资本 | Ant Group、HongShan Group、Zhongguancun Science Park Fund、Bank of China Asset Management 与 Fosun Capital | 股权投资者 | 集中在少数大型、部分带有国资关联的机构 | 任何单一投资者信心撤退,都可能拖住下一轮融资 | 高 | 未披露 | 高 — 未披露资本来源多元化 |
| 通过公开市场进入的战略 / 财务伙伴 | Dongfang Precision (002611.SZ) | 直接持股约 4.36% | 单一具名公开市场股东 | 任何被迫出售或披露出售都可能向市场释放信心下降信号 | 中 | 未披露 | 中 — 不受 Linkerbot 直接控制 |
| 具名客户基础 | Samsung Electronics、Siemens、Stanford、MIT、Cambridge 与 University of Hong Kong | 买方 / 用户 | 公开具名账户数量少;客户总数未披露 | 考虑到集中度未披露,任何单一大型工业账户流失都可能具有重要影响 | 中 | 未披露 | 中 — 未披露客户总数前无法量化 |
| 国际采购 / 贸易目录渠道 | HKTDC, Hannover Messe | 国际买家的发现渠道 | 两个条目均返回未找到结果 | 降低使用这些特定渠道的潜在买家可发现性 | 轻微 | 直销和媒体报道可作为替代渠道 | 低-中 |
| 半导体 / 先进计算供应链 | 未具名零部件供应商 | 关键零部件输入 | 集中度未知;采购未披露 | 美国出口管制可能限制先进部件获取 | 中 | 未披露 | 中 — 取决于美国政策演变 |
各行按严重性排序(最高在前);由于缺少公司特定供应链披露,依赖集中度往往只能定性评估。
[CR020, CR021, CR022, CR023, CR024]制造爬坡、关键人物和资本方集中风险,在发生概率与严重程度的综合评分上最高。
[CR016, CR001, CR025, CR020, CR009, CR019]Linkerbot 依赖高度集中的资本方、未披露的零部件供应链,以及少量公开具名客户。
供应商身份只按类别呈现而非点名,因为已审阅来源均未披露具体零部件供应商。
[CR020, CR021, CR024, CR013]7.4 人员 / 执行与财务 / 模型风险、缓释及终止标准
人员和执行风险集中在唯一公开具名高管身上:2026 年通讯社报道称其为「CEO Alex Zhou」,2025 年报道称其为「Founder and CTO Zhou Yong」,后续独立来源把两个名字指向同一人(「Alex Zhou Yong」);但没有披露董事会、副手或接班计划。以数十亿美元估值看,这是实质治理缺口。公司据报计划在 2026 年底前把研究团队翻倍,同时又设下激进产量增长目标;如果招聘和质量控制能力不能同步扩张,就会产生执行风险,尽管目前没有披露基准人数可供对照。财务 / 模型风险方面,没有来源披露现金、烧钱速度或资金可支撑期限;2026 年 2 月约人民币 15 亿元的 Series B 轮,与 2026 年 5 月约 $3B 的「Series B+」估值没有在任何已审阅来源中被勾稽,稀释和资本效率风险不清楚;没有来源确认当前定价(如约 $920 的 O6 型号)是否支撑可持续毛利。可监控的终止标准 / 论点失效触发项包括:已生效美国采购禁令明确覆盖嵌入式组件;任何确认的产品召回或安全事故;降估值融资、估值下调,或确认未达 2026 年产量目标;唯一具名高管离职且没有继任者;任何独立审计实质性推翻 Linkerbot 的 80%+ 市占率说法。整体看,本章风险画像同时包含印证较充分的行业监管和过剩风险,以及几乎完全缺失的公司层面缓释、认证、保险或应急披露。[CR025, CR026, CR027, CR028, CR029, CR030]
| 角色 / 职能 | 依赖或缺口 | 可能性 | 严重性 | 缓释措施 | 尽调路径 |
|---|---|---|---|---|---|
| CEO / 创始人(Alex Zhou / Zhou Yong) | 所有已审阅报道中唯一具名、面向公众的高管;未确认继任者或副手 | 低-中 | 高 | 未披露 | 索取组织架构图和继任计划 |
| 董事会 / 治理监督 | 未披露董事会构成、独立董事或正式监督结构 | unknown | 中 | 未披露 | 索取股东协议 / 治理文件 |
| 研发 / 工程团队扩张 | 据报道计划到 2026 年底将研究团队翻倍,同时产量目标提升 5-10 倍 | 中 | 中 | 未披露 | 索取招聘计划和当前人数基线 |
| 制造 / 质量控制人员配置 | 未披露匹配快速产能爬坡的员工数或 QC 团队 | 中 | 中 | 未披露 | 索取制造和 QC 人员配置计划 |
各行按严重性排序(最高在前);执行风险相对于已披露增长目标评估,而非基于任何已确认的人员短缺。
[CR025, CR026, CR027, CR028]| 风险 | 可监测触发因素 | 阈值 / 事件 | 行动含义 |
|---|---|---|---|
| 监管(美国采购禁令) | GUARD Act / American Security Robotics Act 最终生效范围 | 明确覆盖非中国品牌机器人内嵌零部件 | 重新评估 Linkerbot 可触达的美国相关需求和估值假设 |
| 运营(可靠性 / 召回) | 产品召回、安全事件或重大可靠性故障的公开报道 | 任何已确认召回或有记录安全事件 | 视为打破投资论点的触发因素,需要立即重新尽调 |
| 财务 / 模型(资本效率) | 降价轮、估值下调,或未达 2026 年产量 / 收入目标 | 任何关于降价轮的公开报道,或确认未达 2026 年 50,000-100,000 台目标 | 重新评估估值立场和资本充足风险评级 |
| 人员 / 执行(关键人) | 公司唯一具名高管离职且未披露继任者 | 任何可信来源报道确认离职 | 视为重大治理和连续性风险事件 |
| 竞争 / 可信度(市场份额说法) | 独立市场份额或出货量审计推翻 Linkerbot 80%+ 说法 | 任何可信第三方审计显示份额实质更低 | 重新评估护城河持久性和估值倍数假设 |
触发因素是可前瞻监测的事件,并非截至报告日期上述结果已经发生的证据。
[CR035, CR036, CR037, CR038, CR034]监管、运营和人员风险各自沿不同路径传导到收入、融资和估值。
[CR001, CR016, CR025, CR030]7.5 要点证据
08估值
8.1 投资论点与反向论点
Linkerbot 的投资论点建立在几项披露之上:高自由度灵巧手品类增长快、成本快速下探,公司声称占有 80%+ 份额;出货增长有交叉印证(2025 年 12 月累计 10,000+ 台,目标 10,000 台 / 月,另有报道称 2026 全年交付 50,000-100,000 台);以及主要机构投资者支持——Ant Group、HongShan Group、Zhongguancun Science Park Fund、Bank of China Asset Management、Fosun Capital——披露估值约 $3B。2026 年 3 月技术大赛获奖和北京市政府认可,为公司自述之外增加了定性可信度。反向论点同样具体:市占率和出货说法由公司自行定义且未经审计,Unitree 与 AgiBot 围绕出货领先的公开争议说明,行业信用问题已经现实存在,并非假设;收入、毛利、现金头寸或人数都未披露;具名客户基础只是一小批 logo 级账户,没有生产还是试点的确认,也没有结果数据;中国国家发改委也公开警示 150 多家国内人形机器人公司存在产能过剩和投机泡沫风险。综合看,当前证据更能支撑增长叙事,而不是尽调级财务或商业结论。[CV001, CV002, CV003, CV037]
| 论点 | 会改变观点的因素 |
|---|---|
| 论点:Linkerbot 领先一个高速增长、成本快速下探且机构背书强的品类 | 独立市场份额 / 出货量审计若确认 80%+ 说法,将大幅强化论点 |
| 论点:出货量增长已有佐证(已出货 10,000+,向 10,000/月爬坡),显示真实商业牵引力 | 若披露的收入 / ARR 数据与出货增长匹配,将确认商业牵引力,而不只是产能牵引力 |
| 反论点:市场份额和出货量说法由公司自定义且未经审计,同业争议(Unitree vs. AgiBot)显示可信度问题已是全行业现实风险 | 独立审计若推翻 80%+ 说法,将确认反论点 |
| 反论点:未披露财务报表、毛利率或现金头寸 | 取得管理账目,或未来 IPO 招股书(如 Unitree)披露,可解决这一问题 |
| 反论点:中国 NDRC 已警告 150+ 家国内人形机器人公司存在全行业产能过剩和泡沫风险 | NDRC 若出台收紧行业的具体政策动作(而不只是表态),将确认反论点;若继续不受约束地增长,则会削弱它 |
论点和反论点直接来自本章及前文已审阅证据;没有进一步尽调前,不预设哪一方占优。
[CV001, CV002, CV003, CV037]8.2 建议、置信度与当前融资背景
考虑到增长叙事强,但经审计财务或独立验证证据几乎完全缺位,给出「继续研究」建议、中等置信度、高风险评级和「偏高」估值立场,比直接买入或回避更站得住;没有来源支持具体目标回报、持有期或退出倍数,因为任何特定投资者的进入价格、股权类别或清算优先权条款都未披露。当前融资背景方面,Linkerbot 在 2026 年 4 月底 / 5 月初左右完成「Series B+」轮,估值约 $3B,并披露下一轮目标约 $6B,但未披露时间或工具。2026 年 2 月约人民币 15 亿元(约 $210M)的 Series B 轮,与约三个月后达到的 $3B 估值,在任何已审阅来源中都没有勾稽,因此该窗口的入场价格纪律难以独立评估。关于实际股权结构,最具体、准独立的证据,是上市公司 Dongfang Precision 通过自身投资者关系披露确认的约 4.36% 直接持股——这是一个真正有用的锚点,但远不足以看清完整股权结构表。整体看,公开证据在方向上符合投资者需求仍在延续的判断,但不能独立确认已披露 $3B 或目标 $6B 是公允价值,而非动量驱动定价。[CV004, CV005, CV006, CV007, CV008, CV009]
| 建议 | 置信度 | 风险评级 | 估值立场 | 决策含义 |
|---|---|---|---|---|
| 继续研究 | 中 | 高 | 偏高 | 在取得市场份额审计、财务披露和融资轮次校准前,不要按已披露的 $3B/$6B 水平投入新资本 |
单行摘要反映本章总体结论;完整推理见随附的论点 / 反论点和情景表。
[CV004, CV040]增长和规模证据很强,但财务与审计证据偏弱,最终给出继续研究建议,而不是买入或回避判断。
[CV001, CV002, CV003, CV004]8.3 牛市 / 基准 / 熊市情景与可比公司组
牛市情景下,Linkerbot 守住 80%+ 份额说法,完成 2026 年 50,000-100,000 台交付目标,并在持续的中国产业政策顺风下,以不低于已披露 $6B 目标的估值完成下一轮融资。基准情景下,增长继续但低于公司表述,下一轮低于 $6B 完成,AgiBot、Zhaowei 和其他独立手部产品进入者带来的价格竞争加剧。熊市情景下,独立审计或竞争对手争议削弱份额说法,国家发改委警示转化为具体政策收紧,制造爬坡不及预期,迫使公司降估值融资或推迟融资;没有来源给出这些情景的显式概率权重,因此这里是定性、以证据为锚的框架,而不是建模预测。对比 Unitree(约 $1.6B,现已提交 IPO,披露达到完整招股书级别)、AgiBot(据报 >RMB 7B,约 $1B,最接近的直接产品品类可比)、Figure AI($39B,西方上限可比)、Apptronik($5B)和 X Square Robot(>$2.8B)这一组可比公司,Linkerbot 披露的约 $3B 估值显著高于最接近的直接手部产品可比公司,但低于最大型西方全栈人形机器人厂商——说明市场给 Linkerbot 的定价更接近广义人形机器人平台,而不是较窄的手部组件供应商。没有来源披露 Linkerbot 或任何可比公司的收入或 EBITDA 倍数,所以这一基准比较依赖绝对轮次估值,而不是标准化倍数;全市场分析师估算(MarketsandMarkets、Goldman Sachs Research)提供宏观顺风背景,但不能直接给 Linkerbot 定价。[CV011, CV012, CV013, CV014, CV015, CV016]
| 情景 | 假设 | 估值 / 回报逻辑 | 关键风险 | 概率信号 |
|---|---|---|---|---|
| 牛市 | 80%+ 份额守住;2026 年 50,000-100,000 台目标达成;下一轮以 $6B 或更高估值完成 | 估值继续重估,逼近或超过已披露的 $6B 目标 | 需要市占率说法经得起审查,且资本继续可得 | 与公司表述和当前融资节奏一致,但未经审计 |
| 基准 | 中等增长延续;下一轮低于 $6B 目标完成;竞争性降价压力加剧 | 估值继续上行,但相对公司自述目标打折 | AgiBot、Zhaowei 等独立灵巧手入局者挤压定价权 | 与竞争组已观察到的商品化趋势一致 |
| 悲观 | 独立审计或竞争对手质疑削弱市占率说法;NDRC 政策收紧;制造爬坡不达预期 | 相对 $6B 目标出现降价轮,或融资明显延迟 | 监管(美国采购禁令)、可信度(市占率争议)和执行(产能缺口)风险叠加 | 与已记录的负面信号一致(NDRC 警示、Unitree-AgiBot 争议、GUARD Act) |
已审阅来源均未给出明确概率权重;「概率信号」反映的是与既有证据的定性一致性,不是模型化概率。
[CV011, CV012, CV013, CV014]| 可比对象 | 指标 | 倍数 / 估值 / 状态 | 可比依据 | 局限 |
|---|---|---|---|---|
| Unitree(中国,全栈人形机器人,自研手部) | 估值 / 申报状态 | ~$1.6B(2025 年 Series C);已递交 IPO,报道称最高募资 $580M | 最接近的中国可比对象,且有公开申报级披露 | 全栈人形机器人厂商,不是 Linkerbot 这样的纯手部组件供应商 |
| AgiBot(中国,全栈人形机器人 + 独立灵巧手销售) | 估值 | 早期轮次报道称估值超过 RMB 7B(~$1B) | 产品品类最接近的直接可比对象(也销售独立灵巧手) | 单一来源、内容偏薄;估值日期和轮次细节未完全交叉验证 |
| Figure AI(美国,全栈人形机器人,自研手部) | 估值 | $39B 投后估值(Series C,募资 >$1B) | 资金充足的西方人形机器人平台,可作上限可比 | 与 Linkerbot 的地域、买方基础和手部采购模式不同 |
| Apptronik(美国,全栈人形机器人,自研手部) | 估值 | $5B(已募资 $520M) | 处于与 Linkerbot 相近扩产阶段的西方中档可比对象 | 地域和商业模式不同(整机 vs. 组件) |
| X Square Robot(中国,具身 AI / 家用机器人) | 估值 | 连续四轮累计 >$2.8B | 融资节奏大体相近的中国可比对象 | 终端市场重心不同(家用 / 消费 vs. Linkerbot 的工业 / 工艺场景) |
估值来自各公司披露或媒体 / 文件报道;这些可比对象(包括 Linkerbot)均未披露足以计算标准化收入或 EBITDA 倍数的财务报表,因此没有使用统一倍数。
[CV015, CV016, CV017, CV018, CV019, CV020]Linkerbot 披露估值显著高于最接近的直接产品品类可比公司 AgiBot,但低于西方最大的全栈人形机器人厂商。
数值取已审阅来源报告的各公司最近披露轮次或备案估值,未按统一日期调整,也未按收入标准化。
[CV015, CV016, CV017, CV018, CV041]Linkerbot 下一轮结果的合理范围从降价轮到达成披露的 $6B 目标;没有来源给出基准情景共识数字。
区间仅作示意,锚定已披露的当前 $3B 估值和 $6B 目标估值,以及本章牛市 / 基准 / 熊市的定性情景逻辑;它们并非来自已披露的公司财务模型。
[CV011, CV012, CV013, CV006]8.4 退出准备度、论点失效触发项与最终尽调问题
已审阅来源没有披露 Linkerbot 自身任何退出事件;行业内最接近退出准备度的证据,是可比同行 Unitree 已提交 STAR Market IPO,说明总体上存在可行退出路径。鉴于 Linkerbot 已披露的规模和机构投资者基础,最终 IPO——可能在中国交易所——合理但目前未获确认;没有来源披露现有股东的 锁定期条款或任何二级转让机制。与估值风险有关的负面信号包括:中国国家发改委 2026 年关于行业产能过剩的警示、Unitree 与 AgiBot 围绕出货领先说法的公开争议(直接类比 Linkerbot 自身未经审计的 80%+ 份额说法),以及美国立法动作(GUARD Act、American Security Robotics Act)可能限制 Linkerbot 可服务的西方市场。四个具体、可监控的论点失效触发项直接来自风险章节:独立审计实质性推翻市占率说法;降估值融资、融资延迟或未达产量目标;已生效美国法律明确覆盖嵌入式组件;公司唯一具名高管离职且没有披露继任者。价值最高的最终尽调问题,是对 Linkerbot 声称的市占率和出货量做独立审计,因为几乎所有其他估值输入都取决于这一说法是否可信;紧随其后的是获取基础财务报表、勾稽 2026 年两次融资事件,以及治理 / 关键人文件。[CV022, CV023, CV024, CV025, CV026, CV027]
| 触发项 | 阈值 | 对投资判断的传导 | 行动含义 |
|---|---|---|---|
| 独立市占率审计 | 任何可信第三方结论显示份额显著低于所称 80%+ | 削弱支撑估值的核心护城河 / 规模论点 | 立即下调建议,并重估估值立场 |
| 融资 / 产量不达预期 | 降价轮、相对 $6B 目标融资明显延迟,或未达 2026 年产量目标 | 暗示商业或财务轨迹弱于披露 | 重估估值立场和资本充足性风险评级 |
| 美国监管行动 | 已生效法律明确覆盖中国制造嵌入式组件(不只是完整机器人) | 压缩 Linkerbot 可触达的西方需求 | 重估非中国收入的市场规模和增长假设 |
| 关键人物离任 | 唯一具名高管确认离任,且未披露继任者 | 移除主要运营 / 战略连续性锚点 | 作为即时治理和连续性风险事件处理 |
触发项是与风险章节已记录风险一致、可前瞻监测的事件;截至报告日,并无证据显示其中任何事件已经发生。
[CV029, CV030, CV031, CV032]| 主题 | 缺失证据 | 重要性 | 负责人 / 尽调路径 |
|---|---|---|---|
| 市占率 / 出货验证 | 对所称 80%+ 高 DoF 份额和单位出货量开展独立审计 | 几乎所有其他估值输入都取决于该说法是否可信 | 委托第三方审计,或开展海关 / 出货数据分析 |
| 财务报表 | 收入、毛利率、现金余额、烧钱速度 | 缺少这些数据,目前只能用可比交易定价,无法采用其他估值方法 | 直接索取管理口径报表,或跟踪未来 IPO 招股书 |
| 融资轮次勾稽 | 勾稽 2026 年 2 月 Series B 轮规模与 2026 年 5 月 Series B+ 估值 | 用于评估两次事件之间的真实稀释和资本效率 | 索取完整股权结构概要,按轮次列明投前 / 投后细节 |
| 治理和关键人物风险 | 董事会构成、继任计划,以及 Alex Zhou / Zhou Yong 身份问题确认 | 关键人物风险集中,且目前未解决 | 索取组织架构图和治理文件 |
尽调问题按「一旦解决,会多大程度改变建议或估值立场」排序。
[CV033, CV034, CV035, CV009]可供 IC 使用的评分显示,市场与规模信号很强,但被偏弱的证据质量和财务透明度得分抵消。
[CV004, CV028, CV040]8.5 要点证据
免责声明
本报告是基于公开证据的尽调快照,不构成投资建议。重要的财务、法律、技术和合同事实仍未公开;任何投资决定前,都应直接向管理层核验,并查阅一手文件。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Linkerbot is a Beijing-based (Haidian District) manufacturer of high-degree-of-freedom dexterous robotic hands for humanoid robots. | 高 | SO009, SO010, SO015 |
| CO002 | Public reporting places Linkerbot's founding in 2023, consistent with the company being described as a 'two-year-old unicorn' in May 2026 wire coverage. | 中 | SO015, SO009 |
| CO003 | A Chinese business-registry snapshot lists the legal entity 灵心巧手(北京)科技股份有限公司 (Linkerbot Beijing Technology Co., Ltd.) with an active ('存续') registration status. | 中 | SO023 |
| CO004 | Linkerbot's core product is the Linker Hand series, spanning roughly 6 to 42 degrees of freedom and multiple actuation types (direct-drive, linkage, tendon-driven). | 高 | SO001, SO002 |
| CO005 | Linkerbot's business model is hardware-first: it sells dexterous hand modules and end effectors to humanoid-robot makers, industrial automation buyers, and research institutions rather than building complete robots itself. | 中 | SO001, SO018 |
| CO006 | As of May 2026 Linkerbot's current stage is a large private, venture-backed company that has just closed a Series B+ round rather than being publicly listed. | 高 | SO009, SO010 |
| CO007 | The company's linkerbot.cn sitemap enumerates at least 13 distinct product pages (L6, L10, L20, L20G, L30, L30Z, O6, O7, D, J, R, W, YCB), indicating a broad hand/gripper product line rather than a single SKU. | 中 | SO019 |
| CO008 | Public reporting names CEO Alex Zhou as Linkerbot's operating leader and spokesperson in May 2026 Reuters-sourced wire coverage. | 高 | SO009, SO010, SO011 |
| CO009 | Earlier Chinese-language coverage from November 2025 names 'Founder and CTO Zhou Yong' as the executive discussing monthly order volume, a different name/title than the 'CEO Alex Zhou' used in May 2026 wire coverage. | 中 | SO008 |
| CO010 | No reviewed source explicitly confirms whether 'Zhou Yong' and 'Alex Zhou' refer to the same individual or whether the executive's title changed from CTO to CEO between November 2025 and May 2026. | 低 | |
| CO011 | No reviewed source discloses an independent board, outside directors, founder equity split, or succession plan for Linkerbot. | 低 | |
| CO012 | Given the technical, customer-facing, and investor-facing centrality of a single named executive across all reviewed coverage, key-person dependence risk is material for Linkerbot. | 中 | SO008, SO009 |
| CO013 | Linkerbot completed a Series A+ financing round reported in late November 2025, its fourth capital injection within an eight-month period at that time. | 中 | SO008 |
| CO014 | Reported Series A+ era investors include Zhejiang Innovation Investment, CDH Investments' Baifu fund, Leju Robotics, and listed companies Dongfang Precision and Aux, alongside earlier backers Ant Group and Sequoia China Seed Fund. | 中 | SO008 |
| CO015 | Linkerbot completed a 'Series B+' financing round that closed in late April/early May 2026 at a valuation of approximately US$3 billion. | 高 | SO009, SO010, SO011, SO013 |
| CO016 | Disclosed Series B+ investors include Alibaba's Ant Group, Sequoia spin-off HongShan Group, the state-backed Zhongguancun Science Park Fund, Bank of China Asset Management, and Fosun Capital. | 高 | SO009, SO010, SO012 |
| CO017 | Linkerbot is reportedly targeting a valuation of about US$6 billion (roughly double the Series B+ level) in its next financing round, without a disclosed timeline or instrument (private round vs. IPO). | 高 | SO009, SO010, SO011 |
| CO018 | No reviewed source discloses Linkerbot's total lifetime capital raised as of May 2026; the only cumulative figure found is $150 million raised since April 2025 as of the December 2025 shipment milestone, which predates the Series B+ round. | 低 | SO001 |
| CO019 | No reviewed source discloses secondary share sales or debt/credit facilities for Linkerbot. | 低 | |
| CO020 | As of May 2026, Linkerbot states it holds over 80% global market share in high-degree-of-freedom (DoF) robotic hands and is the only company mass-producing over 1,000 high-DoF hands per month. | 高 | SO009, SO002, SO025 |
| CO021 | An independent industry tracker corroborates that Linkerbot (灵心巧手) holds 80%+ global share in high-DoF dexterous hands and is the only company delivering more than 1,000 units per month as of early 2026. | 中 | SO025, SO024 |
| CO022 | Linkerbot reports monthly production scaling from almost 5,000 units to a planned 10,000 units 'soon' as of May 2026. | 高 | SO009, SO010, SO011 |
| CO023 | By December 2025, Linkerbot had shipped its 10,000th cumulative dexterous hand and reported a highest single-month shipment above 4,000 units. | 中 | SO001, SO003 |
| CO024 | No reviewed source discloses Linkerbot's revenue, ARR, gross margin, or headcount as of May 2026. | 低 | |
| CO025 | Linkerbot's disclosed named customers include Samsung Electronics and Siemens for industrial/production use, plus stated research use at institutions described as including Stanford and MIT. | 中 | SO001, SO003 |
| CO026 | Linkerbot markets a proprietary software ecosystem, LinkerSkillNet, alongside an 'Open TeleDex' teleoperation system, positioning the company beyond pure hardware into a manipulation-software layer. | 中 | SO001, SO018 |
| CO027 | Linkerbot debuted its full Linker Hand product line and a 'Linker Craftsman' workstation at CES 2026 (January 2026), spanning 11 to 42 degrees of freedom across direct-drive, linkage, and tendon-driven architectures. | 中 | SO002, SO007 |
| CO028 | China's National Development and Reform Commission publicly warned in 2026 that the number of humanoid-robot companies in China has grown past 150 and cautioned against excessive product duplication and speculative bubbles in the sector. | 中 | SO016 |
| CO029 | Linkerbot's HKTDC supplier storefront page and Hannover Messe-style international trade-directory presence could not be independently verified as of the report date; the HKTDC listing returned a page-not-found result. | 低 | SO021 |
| CO030 | An aggregator profile (CB Insights) tracks Linkerbot as a distinct company entity, indicating third-party analyst coverage exists beyond press releases and news wires. | 中 | SO017 |
| CO031 | Linkerbot's official website (linkerbot.cn) is a JavaScript-rendered single-page application, limiting what can be independently extracted from the raw page without a headless browser. | 中 | SO020, SO018 |
| CO032 | Competing Chinese dexterous-hand makers (e.g., AgiBot's OmniHand line) have begun selling standalone hands at aggressive price points, indicating the competitive and pricing environment Linkerbot's milestones sit within. | 中 | SO026 |
| CO033 | No reviewed source discloses Linkerbot's total headcount, number of facilities, or manufacturing footprint as of May 2026. | 低 | |
| CO034 | The May 2026 valuation and market-share claims trace almost entirely to a single Reuters wire story redistributed by BusinessWorld, Straits Times, Yahoo Finance, and The Standard, so the numbers are corroborated across outlets but ultimately share one primary reporting chain plus the company's own statement. | 中 | SO009, SO010, SO011, SO012 |
| CO035 | No reviewed source discloses whether Linkerbot's next targeted US$6 billion valuation would be raised via a private round or a public listing (IPO). | 低 | SO009 |
| CO036 | Sector-level coverage frames Linkerbot as part of a broader 2026 surge of Chinese humanoid-robotics and robotic-component unicorns receiving foreign and domestic institutional capital. | 中 | SO014 |
| CO037 | Linkerbot's reported customer base spans industrial/production deployment (Samsung, Siemens) and academic research use, indicating a dual go-to-market motion rather than a single vertical focus. | 中 | SO001, SO003 |
| CO038 | No reviewed source discloses a specific IPO timetable, listing venue, or underwriter engagement for Linkerbot as of the report date. | 低 | |
| CM001 | The addressable market Linkerbot sells into is best defined as the sub-segment of the humanoid-robotics/embodied-AI hardware stack devoted to dexterous end-effectors (robotic hands and grippers), not the full humanoid-robot market. | 中 | SM006, SM007 |
| CM002 | Excluded from Linkerbot's addressable spend is the far larger bill-of-materials for full humanoid robots (actuators, batteries, locomotion, sensors, AI compute) and adjacent industrial-gripper spend that does not require high-DoF human-like dexterity. | 中 | SM001, SM006 |
| CM003 | Status-quo substitutes for a Linkerbot-style high-DoF dexterous hand include simple 2-3 finger parallel-jaw grippers (e.g., RightHand Robotics) and continued reliance on human manual labor for fine-manipulation tasks. | 中 | SM022, SM023 |
| CM004 | Adjacent markets to Linkerbot's core dexterous-hand product include full humanoid-robot makers that build hands in-house (Unitree, Tesla Optimus) and industrial-automation gripper incumbents that could add DoF over time. | 中 | SM024, SM006 |
| CM005 | MarketsandMarkets estimates the global humanoid robot market at USD 5.41 billion in 2026, growing to USD 50.27 billion by 2035 (CAGR 28.1%). | 中 | SM001 |
| CM006 | Global Market Insights and Future Market Insights both estimate the 2026 global humanoid robot market above USD 10 billion, roughly double MarketsandMarkets' and Grand View Research's estimates for the same year. | 中 | SM002, SM003 |
| CM007 | Goldman Sachs Research frames a wide humanoid-robot TAM range: a base case near $38 billion by 2035 and a 'blue sky' scenario as high as $154 billion if cost and reliability barriers fall faster than expected. | 中 | SM005 |
| CM008 | China Biz Insider reports the narrower China dexterous-hand market alone is projected to exceed $3.7 billion by 2031, a figure closer to a usable SAM proxy for Linkerbot than any whole-humanoid-robot TAM estimate. | 中 | SM006 |
| CM009 | An industry tracker reports Chinese suppliers provide 70-80% of global commercial high-DoF robot hands, and that domestic per-unit pricing has fallen from over $100,000 to below $2,000 in the fastest-moving segment. | 中 | SM007 |
| CM010 | No reviewed source isolates a SOM (serviceable obtainable market) figure specific to Linkerbot; Linkerbot's own 80%+ share claim is applied against an unspecified 'high-DoF dexterous hand' denominator rather than a named third-party sizing study. | 低 | SM018, SM006 |
| CM011 | Reviewed market-sizing estimates for the 2026 whole-humanoid-robot TAM disagree by more than 2x across publishers (roughly $4.2B-$10.9B depending on source), indicating real methodological disagreement rather than a single consensus number. | 中 | SM001, SM002, SM003, SM004 |
| CM012 | Buyers of Linkerbot's dexterous hands span at least three segments: humanoid-robot integrators building their own full robots, industrial-automation end users, and research institutions. | 中 | SM019, SM024 |
| CM013 | Linkerbot's own disclosed named customers (Samsung Electronics, Siemens) sit in the industrial-automation/electronics-manufacturing segment, where the buyer, user, and payer are typically the same corporate procurement function. | 中 | SM019 |
| CM014 | Research-institution buyers (described by Linkerbot as including Stanford and MIT) typically have grant-funded, lab-level budget ownership rather than enterprise procurement budgets, implying smaller per-account revenue but reputational/pipeline value. | 中 | SM019 |
| CM015 | Humanoid-robot integrators (Unitree, Tesla, and others) that build their own hands in-house represent a segment where Linkerbot must sell a 'buy vs. build' value proposition rather than a straightforward component sale. | 中 | SM024, SM007 |
| CM016 | A Morgan Stanley-referenced survey reported in 2026 that many Chinese enterprise buyers want humanoid-robot/dexterous automation but consider current products not yet reliable enough for production deployment, indicating a still-developing adoption trigger for enterprise buyers. | 中 | SM025 |
| CM017 | China's 15th Five-Year Plan (2026-2030) elevates AI-powered robotics, including embodied AI and humanoid robots, to a core national industrial-strategy priority, a growth driver for domestic dexterous-hand suppliers like Linkerbot. | 中 | SM010 |
| CM018 | Falling per-unit hand pricing (from over $100,000 to under $2,000 for comparable-generation Chinese-made hands) is a growth driver that expands the buyer pool beyond well-funded robotics labs to mainstream industrial and even consumer-adjacent buyers. | 中 | SM007 |
| CM019 | Public demonstrations by rival humanoid-robot makers (e.g., televised performances and public marathon events referenced in 2026 wire coverage) are cited as a driver of investor and buyer interest across the whole China humanoid-robotics sector, indirectly benefiting hand suppliers like Linkerbot. | 中 | SM020 |
| CM020 | China's National Development and Reform Commission publicly warned in 2026 that more than 150 humanoid-robot companies now exist domestically and cautioned against excessive duplication and speculative bubbles, an adoption/valuation-discipline constraint for the whole sector. | 中 | SM011 |
| CM021 | US lawmakers introduced bipartisan bills in March 2026 (the American Security Robotics Act and the GUARD Act) that would ban US federal government procurement and use of humanoid and ground robots made by Chinese firms, a direct adoption constraint for Chinese-made components including dexterous hands in the US public sector. | 高 | SM012, SM013, SM027 |
| CM022 | US export-control tightening on advanced semiconductors relevant to robotics compute is a further adoption/supply constraint that could affect the performance ceiling of Chinese-made dexterous-hand electronics sold into Western-adjacent supply chains. | 中 | SM014 |
| CM023 | International safety standards for industrial and collaborative robots (ISO 10218, ISO/TS 15066) create a compliance/certification hurdle that dexterous-hand suppliers must clear before large industrial buyers will deploy hands at scale. | 中 | SM016, SM017 |
| CM024 | The EU's AI Act regulatory framework is a further compliance constraint that could slow adoption timelines for embodied-AI hardware, including dexterous hands, sold into European enterprise or public-sector buyers. | 低 | SM028 |
| CM025 | Capital intensity of scaling hand manufacturing from thousands to tens of thousands of units per month is itself an adoption constraint, since supply reliability (not just unit cost) determines whether large industrial buyers will commit to a single supplier. | 中 | SM018, SM007 |
| CM026 | No reviewed source reconciles the more-than-2x spread in 2026 whole-market TAM estimates or explains why Grand View Research's and MarketsandMarkets' numbers are roughly half of Global Market Insights' and Future Market Insights' estimates for the same year. | 中 | SM001, SM002 |
| CM027 | No reviewed source provides an independently audited SAM or SOM figure specific to high-DoF dexterous hands (as distinct from whole humanoid robots or generic industrial grippers); the closest proxy is China Biz Insider's $3.7B-by-2031 China dexterous-hand estimate. | 低 | SM006 |
| CM028 | CNBC reported in June 2026 on a Morgan Stanley forecast specifically addressing the China humanoid robot market, indicating continued sell-side analyst attention to China-specific sizing distinct from global totals. | 中 | SM026 |
| CM029 | IDC's 2026 commentary on humanoid-robotics commercialization frames the market as still pre-mass-commercialization, with production and reliability, not demand, as the binding constraint in the near term. | 中 | SM009 |
| CM030 | Linkerbot's own claimed 80%+ share of the 'global high-DoF dexterous hand market' is a company-defined category boundary that does not match any of the whole-humanoid-robot TAM figures published by MarketsandMarkets, Grand View Research, Global Market Insights, or Future Market Insights. | 中 | SM018, SM001 |
| CM031 | The adoption path for an industrial buyer typically runs from pilot/evaluation units, through reliability and safety certification, to production-line integration and multi-unit reorder, a path that is only partially complete for most humanoid-robot and dexterous-hand deployments as of 2026 per Morgan Stanley-referenced survey data. | 中 | SM025, SM019 |
| CM032 | Research-institution adoption (labs at Stanford/MIT per company statements) functions as an early low-revenue, high-visibility adoption channel that can accelerate enterprise trust rather than a primary revenue segment. | 中 | SM019 |
| CM033 | No reviewed source separates Linkerbot's revenue or unit sales by buyer segment (industrial vs. research vs. robot-integrator), so segment-level revenue mix cannot currently be estimated. | 低 | |
| CM034 | Sector-level coverage frames 2026 as a peak year for Chinese humanoid-robotics and robotic-component fundraising, which is itself a demand-side growth driver via well-capitalized robot-integrator customers able to buy hands at scale. | 中 | SM021 |
| CM035 | US Treasury OFAC sanctions authority is a latent adoption-constraint mechanism (not currently applied specifically to Linkerbot in any reviewed source) that could be extended to Chinese robotics-component makers if national-security concerns escalate. | 低 | SM015 |
| CM036 | No reviewed source models price elasticity of demand for dexterous hands, i.e., how much unit volume would grow if prices fell further from the already-disclosed sub-$1,000 tier toward the sub-$100 tier some Chinese competitors have publicly projected within three years. | 低 | |
| CP001 | The competitive landscape for Linkerbot spans direct dexterous-hand specialists (Inspire Robots, Shadow Robot, Zhaowei Mechatronics), full-stack humanoid-robot makers that build their own hands (Unitree, AgiBot, Fourier Intelligence, UBTech, Tesla), and non-China humanoid-robot makers building in-house hands (Figure AI, Apptronik, 1X Technologies). | 中 | SP001, SP003 |
| CP002 | Status-quo substitutes for buyers that do not need full human-like dexterity include simple parallel-jaw grippers (e.g., RightHand Robotics) and continued manual human labor. | 中 | SP022 |
| CP003 | Research-oriented dexterous hands such as Wonik Robotics' Allegro Hand and the open-source LEAP Hand represent a lower-cost, lab-oriented adjacent category rather than mass-production industrial competitors. | 中 | SP023, SP024 |
| CP004 | The most credible internal-build threat comes from humanoid-robot integrators themselves: AgiBot, Unitree, Fourier Intelligence, and UBTech all design and manufacture their own hands rather than buying from independent suppliers like Linkerbot. | 中 | SP005, SP008, SP010 |
| CP005 | Unitree disclosed three tiers of self-developed dexterous hands in its March 2026 IPO prospectus: the five-finger Dex5-1 (20 DoF, 94 tactile sensors, 4.5kg payload, ±1mm precision, 1kg weight), the three-finger Dex3-1 (7 DoF, 33 sensors), and the Dex1-1 gripper (120N grip force, 20kg payload). | 中 | SP001 |
| CP006 | Unitree was valued at approximately $1.6 billion after a Series C round reported in July 2025 and later filed for an IPO reported at up to $580 million in proceeds as humanoid-robot sales began to surpass its legacy quadruped-robot business. | 中 | SP006, SP007 |
| CP007 | Shadow Robot's dexterous hand is described as the acknowledged technical benchmark at 24 DoF and 100+ sensors, but at over $140,000 per hand its volume production has stalled, illustrating a high-end technology leader that cannot compete on Linkerbot's cost/scale axis. | 中 | SP001, SP026 |
| CP008 | AgiBot began direct standalone sales of its OmniHand series, pricing a 10-degree-of-freedom model at RMB 9,800 (about US$1,350), breaking from the industry norm of bundling hands only with complete humanoid robots. | 中 | SP002, SP012 |
| CP009 | Zhaowei Mechatronics offers at least two dexterous-hand product lines: the DM17 (linear-motor direct drive, 17 active DoF, 5-12N fingertip force) and the LM06 (linkage transmission, 6 active/16 total DoF, 150N grip force), entering the category from a component-manufacturing background. | 中 | SP001 |
| CP010 | Tesla's Optimus Gen3 hand is estimated (via teardown-based analyst reads) at 17 active plus 5 passive DoF with 22 Hall position sensors, six-axis force sensing, and 94 tactile contact points, and Elon Musk has stated hand R&D represents roughly half of Optimus's total engineering effort, with a target cost of $20,000-$25,000 per robot at 1 million annual production. | 中 | SP001 |
| CP011 | Figure AI, a US humanoid-robot maker that designs its own hands, raised more than $1 billion in a Series C round at a $39 billion post-money valuation. | 中 | SP013 |
| CP012 | Apptronik, another US humanoid-robot maker, raised $520 million at a $5 billion valuation, explicitly positioning itself to reach market ahead of Tesla. | 中 | SP014 |
| CP013 | 1X Technologies raised $100 million in a Series B round, a smaller capital base than the largest China- and Figure-AI-scale competitors, indicating a more fragmented funding tier among Western humanoid-robot makers. | 中 | SP015 |
| CP014 | X Square Robot, a Chinese embodied-AI competitor, completed four consecutive financing rounds and surpassed a US$2.8 billion valuation while explicitly targeting household/consumer manipulation tasks rather than Linkerbot's stated focus on high-value craftsmanship skills. | 中 | SP016, SP017 |
| CP015 | UBTech's Walker S2 humanoid robot began mass production and delivery with disclosed orders exceeding 800 million yuan, indicating a full-stack competitor with meaningful commercial order backlog of its own. | 中 | SP011, SP010 |
| CP016 | Galbot raised RMB 2.5 billion in March 2026 and Robotera raised over US$200 million in May 2026, both signaling continued heavy capital inflow into China-based humanoid-robot and embodied-AI makers that could eventually build or buy dexterous hands at scale. | 中 | SP019, SP021 |
| CP017 | On pricing, Linkerbot's own O6 model is reported at approximately ¥6,666 (about $920), positioned below AgiBot's ¥9,800 (~$1,350) 10-DoF OmniHand and far below Shadow Robot's >¥1,000,000 (~$140,000) benchmark hand and generic overseas comparables at ¥1.1M-1.6M ($150K-$220K). | 中 | SP001, SP002 |
| CP018 | No reviewed source discloses standardized list pricing for Unitree's, Fourier's, or UBTech's dexterous hands sold as standalone components, because these companies bundle hands into complete humanoid-robot sales rather than pricing them separately. | 低 | SP005, SP008, SP010 |
| CP019 | Linkerbot's disclosed go-to-market channel is direct international press/trade-show visibility (CES 2026, PR Newswire) plus direct sales to named industrial and research accounts, rather than a disclosed reseller or distributor network. | 中 | SP030 |
| CP020 | No reviewed source discloses formal regulatory or safety certifications held by Linkerbot, Inspire Robots, AgiBot, or Zhaowei specific to their dexterous-hand products, unlike the broader ISO industrial-robot safety standards referenced at the sector level. | 低 | |
| CP021 | Switching cost for a robot integrator that has already engineered its control stack around one supplier's hand (mounting, wiring harness, control API, and manipulation-skill data) is likely to be non-trivial, creating some lock-in once a hand is designed into a robot platform. | 中 | SP001 |
| CP022 | Multi-homing (a robot integrator qualifying more than one hand supplier) appears plausible given that Linkerbot's own product line spans 13+ SKUs at different price/performance tiers, but no source confirms any single integrator currently dual-sources hands from Linkerbot and a rival. | 低 | |
| CP023 | Distribution power sits disproportionately with the full-stack humanoid-robot makers (Unitree, AgiBot, Figure AI, UBTech) because they control the end customer relationship; independent hand suppliers like Linkerbot depend on being designed into someone else's robot or selling direct to research/industrial accounts. | 中 | SP005, SP010 |
| CP024 | Linkerbot's claimed moat is production scale and cost leadership (80%+ share, sub-$1,000 pricing, 1,000+ units/month) rather than a single technical breakthrough, since Shadow Robot already holds a higher-DoF technical benchmark at far higher cost. | 中 | SP001, SP003, SP030 |
| CP025 | Commoditization risk is material: AgiBot's move to sell a 10-DoF hand at ~$1,350 standalone, and Zhaowei's entry from a generic component-manufacturing background, both signal that dexterous hands could become a price-competitive commodity category rather than a differentiated one. | 中 | SP002, SP001 |
| CP026 | An adverse public dispute exists over shipment-volume leadership: Humanoids Daily reported Unitree publicly rebutting AgiBot's claim to the global No. 1 humanoid-shipment title, indicating that competitor self-reported scale claims (including Linkerbot's own 80%+ share claim) should be treated with caution absent independent audit. | 中 | SP028 |
| CP027 | China's National Development and Reform Commission publicly warned in 2026 that more than 150 humanoid-robot companies now exist domestically, an adverse structural signal that the whole competitive set (Linkerbot included) faces overcapacity and potential consolidation risk. | 中 | SP029 |
| CP028 | No reviewed source discloses a direct head-to-head win/loss record between Linkerbot and any single named competitor for a specific customer account. | 低 | |
| CP029 | Fourier Intelligence's humanoid robots (GR-1, GR-2) are positioned across research, rehabilitation, and service-robotics use cases, a different buyer mix than Linkerbot's stated industrial/craftsmanship focus, reducing direct overlap even though Fourier builds its own hands. | 中 | SP008, SP009, SP027 |
| CP030 | No reviewed source discloses whether any Western humanoid-robot maker (Figure AI, Apptronik, 1X Technologies) currently sources any hand components from Chinese suppliers, an open question relevant to potential future export-control exposure. | 低 | |
| CP031 | Inspire Robots (因时机器人) is frequently cited alongside Linkerbot and Unitree as one of the fastest-scaling Chinese dexterous-hand suppliers, though its official site could not be extracted beyond the page title during this review. | 中 | SP004, SP003 |
| CP032 | XHand1, listed on an independent humanoid-robot product directory, represents a further Chinese dexterous-hand entrant beyond the companies with dedicated funding-round press coverage reviewed for this chapter. | 低 | SP025 |
| CP033 | No reviewed source provides a standardized, independently verified feature/capability comparison across all named competitors; the feature matrix in this chapter is assembled from separate company and press disclosures with different testing methodologies and should be treated as directional, not audited. | 中 | SP001 |
| CP034 | Robotera (星动纪元) markets itself as the industry's only company with full self-developed hardware-and-software embodied-AI stack, a differentiation claim that, if accurate, would reduce its reliance on third-party dexterous-hand suppliers including Linkerbot. | 低 | SP020 |
| CP035 | No reviewed source discloses Linkerbot's win rate, deal pipeline, or competitive displacement instances against any specific named rival, leaving competitive-moat durability primarily inferable rather than directly evidenced. | 低 | |
| CI001 | Linkerbot's disclosed revenue stream is hardware sales of the Linker Hand product family (13+ SKUs spanning 6-42 DoF), sold directly to industrial, humanoid-robot-integrator, and research buyers. | 中 | SI005, SI006 |
| CI002 | Linkerbot also markets a software/platform layer (LinkerSkillNet manipulation-skill ecosystem, Open TeleDex teleoperation system) alongside hardware, but no source discloses whether this generates separate license, subscription, or data-access revenue distinct from hardware sales. | 低 | SI005 |
| CI003 | An independent industry tracker reports Linkerbot's O6 model list price at approximately ¥6,666 (about $920), the only specific unit price publicly attributed to Linkerbot found in reviewed sources. | 中 | SI011 |
| CI004 | A company co-founder is quoted predicting Linkerbot hand prices could fall below ¥500 (about $70) within three years, implying an aggressive planned pricing/cost trajectory rather than a stable list price. | 中 | SI011 |
| CI005 | No reviewed source discloses Linkerbot's revenue mix across hardware tiers, software, or services, so revenue-recognition treatment (point-in-time hardware sale vs. any recurring software/data revenue) cannot currently be assessed. | 低 | |
| CI006 | Linkerbot's disclosed go-to-market motion is direct sales to named industrial and research accounts plus international press/trade-show visibility (CES 2026), not a disclosed channel-partner or reseller network. | 中 | SI006, SI005 |
| CI007 | No reviewed source discloses sales-cycle length, CAC, payback period, or any channel-economics proxy for Linkerbot's go-to-market motion. | 低 | |
| CI008 | No reviewed source discloses Linkerbot's manufacturing cost per unit, bill-of-materials breakdown, gross margin, or gross-margin drivers. | 低 | |
| CI009 | An industry tracker attributes Linkerbot's low list pricing partly to proprietary components (PEEK engineering polymer, in-house micro-reducers) and manufacturing scale, implying cost drivers are componentry and volume rather than disclosed labor or overhead figures. | 中 | SI012, SI006 |
| CI010 | No reviewed source discloses Linkerbot's working-capital position, inventory levels, or capital-expenditure plans for manufacturing capacity expansion beyond the general statement that new capital funds 'production capacity' improvements. | 低 | SI001 |
| CI011 | No reviewed source discloses service-delivery costs (support, warranty, integration services) associated with Linkerbot's hardware sales. | 低 | |
| CI012 | Linkerbot's disclosed public traction metrics are unit-shipment and production-scale figures (10,000+ cumulative hands shipped by December 2025; ~5,000 to a targeted 10,000 units/month by mid-2026) rather than financial metrics. | 高 | SI005, SI008, SI009 |
| CI013 | No reviewed source discloses Linkerbot's revenue, ARR, gross merchandise value, active-customer count, or utilization metrics, leaving a complete gap between public unit-shipment traction and any financial-statement-level metric. | 低 | |
| CI014 | A Chinese-language funding report states Linkerbot targeted delivering 50,000 to 100,000 dexterous-hand units in 2026, a target roughly 5-10x its disclosed December 2025 cumulative shipment level. | 中 | SI001 |
| CI015 | Linkerbot completed a Series B round of approximately RMB 1.5 billion (roughly $210 million) reported on 2026-02-12, led by Daode Investment and Shengshi Investment, with participation from Yingyuan Puyuan Fund, Nova Star Cloud, Xinding Capital, Zhonghe Capital, Zhilu Capital, Chengzhu Investment, Linyun Capital, Jiaming Haochun, Caixin Capital, and Singapore Eastern Epic Capital. | 中 | SI001 |
| CI016 | Disclosed use of proceeds for the February 2026 Series B round is core product/technology R&D, production-capacity expansion, and building 'full-stack foundation' manipulation capability, rather than working-capital or debt-service needs. | 中 | SI001 |
| CI017 | Listed company Dongfang Precision (东方精工, 002611.SZ) discloses holding an approximately 4.36% direct equity stake in Linkerbot, confirmed unchanged in a June 2026 investor-relations response referencing the company's 2025 annual report. | 中 | SI002, SI003 |
| CI018 | By early May 2026 Linkerbot had closed a further 'Series B+' round at approximately a $3 billion valuation, with named investors including Ant Group, HongShan Group, Zhongguancun Science Park Fund, Bank of China Asset Management, and Fosun Capital, and disclosed a target of roughly $6 billion for its next round. | 高 | SI008, SI009, SI019 |
| CI019 | No reviewed source discloses Linkerbot's cash on hand, monthly burn rate, or runway in months as of the report date. | 低 | |
| CI020 | No reviewed source discloses any debt or project-finance obligations for Linkerbot; all disclosed capital appears to be primary equity financing. | 低 | |
| CI021 | No reviewed source discloses an explicit 'next-round trigger' (e.g., a specific shipment, revenue, or production milestone) that Linkerbot has said must be hit before pursuing the disclosed $6 billion target valuation. | 低 | SI008 |
| CI022 | China's National Development and Reform Commission publicly warned in 2026 that more than 150 humanoid-robot companies now exist domestically and cautioned against speculative bubbles, an adverse macro signal relevant to how aggressively Linkerbot's rapid financing cadence and rising valuation should be read. | 中 | SI014 |
| CI023 | A comparable-company filing (Unitree's March 2026 STAR Market IPO prospectus) illustrates that at least one peer humanoid/dexterous-hand manufacturer has reached full audited-financial-statement disclosure via a public listing, a disclosure standard Linkerbot has not yet reached as a private company. | 中 | SI004 |
| CI024 | Manufacturing scale-up in this sector is capital-intensive: peer coverage of AgiBot's push past a 10,000-unit production threshold frames comparable capital and operational requirements for reaching Linkerbot's own stated 10,000-unit/month target. | 中 | SI026 |
| CI025 | No reviewed source discloses Linkerbot's per-unit gross margin, so it cannot currently be determined whether the disclosed ~$920 O6 list price and the co-founder's stated sub-$70 three-year target price are consistent with a sustainable gross margin or reflect below-cost, growth-stage pricing. | 低 | SI011 |
| CI026 | No reviewed source discloses Linkerbot's customer acquisition cost, contract length, warranty cost, or return/defect rate, all of which are necessary inputs to a genuine unit-economics model. | 低 | |
| CI027 | No reviewed source discloses Linkerbot's headcount, so per-employee revenue or R&D-spend proxies cannot be estimated. | 低 | |
| CI028 | CB Insights maintains an analyst-tracked company profile for Linkerbot, indicating some independent financial-data-provider coverage exists beyond press releases, though the specific financial fields available were not disclosed in the reviewed excerpt. | 中 | SI013 |
| CI029 | Whole-market analyst estimates (MarketsandMarkets, Goldman Sachs Research) for the broader humanoid-robotics sector suggest a capital-intensive, still-early-commercialization industry context, consistent with Linkerbot's own reliance on repeated equity financing rather than disclosed self-funding from operations. | 中 | SI022, SI023 |
| CI030 | No reviewed source reconciles the ~RMB 1.5 billion February 2026 Series B figure with the ~$3 billion valuation reported for the 'Series B+' round roughly three months later, so the amount of capital raised versus valuation step-up between the two rounds cannot be fully decomposed from public sources. | 中 | SI001, SI008 |
| CI031 | Total lifetime capital raised by Linkerbot as of the report date is not disclosed in any single reviewed source; the only fully cumulative figure available (excluding the Series B+ round) is $150 million raised since April 2025 as of the December 2025 shipment milestone. | 低 | SI005 |
| CI032 | The overall financial-diligence picture for Linkerbot combines strong, multiply-corroborated unit-shipment and financing-round evidence with a near-total absence of revenue, margin, cost-structure, and balance-sheet disclosure, typical of a fast-scaling private hardware company but a material blocker to any independent financial valuation. | 中 | SI005, SI008, SI017, SI002 |
| CI033 | Given the capital-intensive nature of scaling hardware manufacturing (illustrated by peer capital raises and production-threshold coverage), Linkerbot's financing dependency on continued equity rounds is likely to persist until (or unless) hardware gross margins and/or a software/data revenue layer mature enough to fund growth internally. | 中 | SI026, SI001 |
| CI034 | No reviewed source discloses a pathway or timeline for Linkerbot to reach the audited-financial-statement disclosure standard illustrated by comparable peer Unitree's IPO prospectus, whether via its own eventual listing or otherwise. | 低 | SI004 |
| CI035 | China Economic Net's coverage of the broader dexterous-hand sector frames 2025-2026 as a period of rapid commercialization race dynamics across many funded entrants, a context in which Linkerbot's specific unit economics remain undisclosed relative to peers. | 中 | SI021 |
| CI036 | No reviewed source discloses Linkerbot's pricing for its higher-DoF flagship models (e.g., L30, L20G) beyond the O6 entry-tier price, so a full pricing ladder across the 13+ SKU product line cannot currently be constructed. | 低 | SI011 |
| CE001 | Linkerbot's product, in customer workflow terms, is a dexterous robotic hand end-effector that a customer mounts on a robot arm or humanoid platform and controls via a ROS/ROS2 driver stack and Python/C++ SDK rather than a complete robot. | 中 | SE002, SE003 |
| CE002 | The customer integration workflow documented in the ROS2 SDK repository is: install the SDK on Ubuntu 22.04 with ROS Humble and Python 3.10+, configure hand parameters via a setting.yaml file, then drive the hand through ROS topics or the LinkerHandApi Python class. | 中 | SE003 |
| CE003 | Linkerbot's official sitemap lists at least 13 distinct product pages (models L6, L10, L20, L20G, L30, L30Z, O6, O7, D, J, R, W, YCB), indicating a broad product-line map beyond a single flagship SKU. | 中 | SE008 |
| CE004 | Linkerbot's ROS2 SDK README documents hardware support for a different, partially overlapping model list (O6, L6, L7, O7, L10, L20, G20, L21) than the official sitemap's product pages, indicating the developer-facing SDK does not yet cover the full commercial product catalog (or uses different internal naming for some models, e.g., G20 vs. L20G). | 中 | SE003, SE008 |
| CE005 | The flagship L30 model is described in company materials as achieving approximately ±0.2mm repeatability, positioned for precision-assembly workflows. | 中 | SE006 |
| CE006 | The O6 model is described in company materials as weighing 370g while delivering up to 50kg grip payload, achieved via PEEK engineering polymer and in-house-developed micro-reducers. | 高 | SE007, SE013 |
| CE007 | Linkerbot's dexterous hands cover three distinct actuation architectures — direct-drive, linkage, and tendon-driven — within a single product family, which an independent industry tracker highlights as unusual full technology-route coverage compared to most single-architecture rivals. | 高 | SE014, SE007 |
| CE008 | The industrial-tier Linker Hand L20 is described in company materials as using a proprietary 'Super Electric Cylinder' achieving 90% drive efficiency, 200N screw-end thrust, and a tested lifespan exceeding one million cycles. | 中 | SE007, SE013 |
| CE009 | Linkerbot markets a multi-modal 'sensing skin' tactile-sensing layer across its hand family, extended through the 'Open TeleDex' teleoperation system and the 'LinkerSkillNet' manipulation-skill software ecosystem. | 中 | SE006, SE011 |
| CE010 | No GitHub repository or documentation page under the linker-bot organization was found by name for 'Open TeleDex' or 'LinkerSkillNet'; the closest matching repository is a general teleoperation-arm utility, suggesting these may be product/marketing names layered over more narrowly named underlying code modules, or are not yet open-sourced. | 中 | SE001 |
| CE011 | Linkerbot provides a URDF (Unified Robot Description Format) model repository for its hands, indicating support for standard robotics simulation and motion-planning tooling (e.g., Gazebo, MoveIt) rather than a fully closed integration stack. | 中 | SE004 |
| CE012 | The ROS2 SDK is documented as supporting Ubuntu 22.04 and ROS Humble specifically, a relatively current but not bleeding-edge robotics software stack as of the report date, implying integration is realistic for most contemporary robotics labs and integrators. | 中 | SE003 |
| CE013 | No reviewed source discloses a formal customer support SLA, warranty terms, or dedicated integration-support program for Linkerbot's hardware beyond a general FAQ page on the company's JavaScript-rendered website. | 低 | SE026 |
| CE014 | Linkerbot publicly reveals product roadmap milestones through trade-show events rather than a published roadmap document: the CES 2026 debut introduced the full Linker Hand line plus the 'Linker Craftsman' workstation spanning 11 to 42 DoF across direct-drive, linkage, and tendon-driven architectures. | 中 | SE007, SE013, SE019 |
| CE015 | No reviewed source discloses a forward-looking public roadmap (e.g., named next-generation models or target release dates) beyond the CES 2026 product reveal itself. | 低 | |
| CE016 | Linkerbot's stated technology differentiation rests on covering all three major actuation approaches (direct-drive, linkage, tendon) in one product family, proprietary materials (PEEK polymer) and micro-reducer manufacturing, and a claimed 80%+ market share built on cost/scale leadership rather than a single patentable breakthrough. | 中 | SE014, SE007 |
| CE017 | No reviewed source discloses Linkerbot's patent portfolio, patent filings, or any granted intellectual-property protection specific to its actuation, sensing, or manufacturing methods. | 低 | |
| CE018 | No reviewed source discloses Linkerbot's manufacturing facility count, location(s) beyond its Beijing headquarters, or supply-chain/component-sourcing detail for critical parts (e.g., motors, tactile sensors, PEEK polymer components). | 低 | |
| CE019 | No reviewed source discloses any formal safety, quality-management, or industrial-robot compliance certification (e.g., against ISO 10218 or ISO/TS 15066) specifically held by Linkerbot for its hand products. | 低 | SE016, SE017 |
| CE020 | No reviewed source discloses a data-privacy or security policy specific to manipulation-skill data collected via LinkerSkillNet or teleoperation sessions run through Open TeleDex. | 低 | |
| CE021 | China's National Development and Reform Commission publicly warned in 2026 of excessive product duplication and quality/safety risk implied by more than 150 domestic humanoid-robot companies, a sector-level adverse signal relevant to assessing whether rapid product-line expansion (13+ Linkerbot SKUs) has kept pace with quality and safety controls. | 中 | SE025 |
| CE022 | Linkerbot's own product materials note the ROS SDK README includes 'safety notes' for hardware operation, indicating at least an informal, developer-facing safety-guidance layer even without a named third-party certification. | 中 | SE002 |
| CE023 | Linkerbot's disclosed named customers/users span industrial production (Samsung, Siemens per company statements) and research institutions (Stanford, MIT), indicating the product has been deployed (or at least piloted) beyond internal demonstration settings. | 中 | SE006, SE015 |
| CE024 | No reviewed source discloses field reliability data (mean time between failures, defect rate, or return rate) for any Linkerbot hand model in production deployment. | 低 | |
| CE025 | An independent competitor-product listing exists for AgiBot's OmniHand, providing a useful architecture comparison point: OmniHand's standalone commercial hand is disclosed at 10 DoF, a lower specification than Linkerbot's flagship L-series and O-series tiers. | 中 | SE023 |
| CE026 | CB Insights maintains an analyst-tracked profile for Linkerbot, indicating the product and company are tracked by at least one independent data provider beyond company and press materials. | 中 | SE024 |
| CE027 | No reviewed source discloses a unified architecture diagram showing how LinkerSkillNet, Open TeleDex, the ROS/ROS2 SDK, and hardware firmware relate to one another as a single stack; the architecture in this chapter is assembled from separate product and developer-facing disclosures. | 中 | SE006, SE003 |
| CE028 | Linkerbot's GitHub organization hosts multiple SDK variants (Python, C++, ROS1, ROS2) plus a simulator and URDF repository, indicating a broader developer-tooling investment than a minimal single-language driver would represent. | 中 | SE001, SE002, SE003, SE004 |
| CE029 | No reviewed source discloses adoption metrics (stars, forks, external contributors, or download counts) for Linkerbot's GitHub repositories, so developer-community traction cannot be quantified from the reviewed excerpts. | 低 | SE001 |
| CE030 | No reviewed source documents a specific production-line integration case study (e.g., a named factory deployment with measured before/after throughput) for any Linkerbot customer. | 低 | |
| CE031 | Linkerbot's product-line breadth (13+ SKUs spanning 6-42 DoF at different actuation types) is itself a form of differentiation versus rivals that typically offer a narrower single-architecture line, per an independent industry tracker. | 中 | SE014 |
| CE032 | No reviewed source discloses quality-control processes (incoming inspection, in-line testing, outgoing QA) used in Linkerbot's manufacturing operations. | 低 | |
| CE033 | China Economic Net's coverage frames the broader dexterous-hand sector's 2025-2026 commercialization race as prioritizing shipment volume, which is a relevant context for assessing whether Linkerbot's rapid SKU proliferation has been matched by proportional quality-assurance investment, an unconfirmed but material question. | 低 | SE022 |
| CE034 | No reviewed source discloses whether Linkerbot's teleoperation (Open TeleDex) sessions or the underlying manipulation datasets are used to train third-party AI models beyond Linkerbot's own LinkerSkillNet, an open question relevant to data-rights and privacy diligence. | 低 | |
| CE035 | The gap between a demonstrably active open-source developer-tooling investment (multiple SDKs, simulator, URDF models) and a near-total absence of disclosed certifications, reliability data, or manufacturing-quality documentation suggests Linkerbot's public disclosure strategy currently favors developer adoption and press visibility over formal enterprise-trust documentation. | 中 | SE001, SE019 |
| CU001 | Linkerbot's disclosed customer base segments into at least three verticals: industrial/electronics manufacturing (Samsung Electronics, Siemens), research institutions (Stanford, MIT, Cambridge, University of Hong Kong per company and press statements), and humanoid-robot integrators as an implied buyer class. | 高 | SU002, SU001, SU004 |
| CU002 | For industrial customers, the buyer, user, and payer are typically the same corporate procurement/automation-engineering function; for research customers, the payer is grant/lab funding rather than the using researcher directly. | 中 | SU002 |
| CU003 | No reviewed source discloses Linkerbot's customer base by geography beyond implied China-domestic manufacturing plus named international research institutions (Stanford, MIT, Cambridge) and one Korean industrial customer (Samsung Electronics). | 中 | SU002, SU001 |
| CU004 | No reviewed source discloses customer segmentation by company size (SMB vs. enterprise) or by revenue band per account. | 低 | |
| CU005 | Linkerbot's disclosed adoption trajectory shows monthly orders surpassing 1,000 units by November 2025, cumulative shipments exceeding 10,000 units by December 2025, and a targeted scale-up to 10,000 units per month by mid-2026 from almost 5,000. | 高 | SU004, SU002, SU007, SU008 |
| CU006 | A separate Chinese-language report states Linkerbot targeted delivering 50,000 to 100,000 dexterous-hand units across all of 2026, a target substantially above its disclosed end-2025 cumulative shipment level. | 中 | SU014 |
| CU007 | No reviewed source discloses the number of distinct active customer accounts, deployment locations, or utilization rates for Linkerbot's shipped hands, so unit-shipment counts cannot be converted into an active-account or utilization metric. | 低 | |
| CU008 | Linkerbot won the championship title at the 9th Zhongguancun International Advanced Technology Competition (announced March 26, 2026) among more than 2,100 participating global projects, a form of third-party validation distinct from a direct sales metric. | 中 | SU005 |
| CU009 | Samsung Electronics is named as an industrial customer of Linkerbot across multiple independent sources (China Biz Insider, PR Newswire, Interesting Engineering), providing multi-source corroboration for this specific named account. | 高 | SU004, SU002, SU001 |
| CU010 | Siemens is named as an industrial customer of Linkerbot in both a company press release and independent China Biz Insider reporting, though no source specifies whether the deployment is in production or pilot phase. | 中 | SU002, SU004 |
| CU011 | Stanford University is named as a research user of Linkerbot's hands in both company materials and independent press coverage (Interesting Engineering), though no source specifies a named lab, publication, or specific research outcome. | 中 | SU002, SU001 |
| CU012 | MIT and Cambridge are named as research users of Linkerbot's hands in a company press release, without independent third-party corroboration beyond the company's own statement in the sources reviewed for this chapter. | 中 | SU002 |
| CU013 | The University of Hong Kong is named as a client of Linkerbot by an independent press source (Interesting Engineering), a named customer not mentioned in Linkerbot's own reviewed press releases. | 中 | SU001 |
| CU014 | No reviewed source distinguishes, for any named customer (Samsung, Siemens, Stanford, MIT, Cambridge, University of Hong Kong), whether the deployment is a production installation, a pilot, or a one-off demonstration/loan of equipment. | 低 | |
| CU015 | All named-customer evidence found for this chapter is dated between November 2025 and May 2026, meeting the freshness bar for a mid-2026 report but with no source confirming these relationships are still active as of the report date itself. | 中 | SU001, SU002, SU004 |
| CU016 | No reviewed source discloses net revenue retention, gross revenue retention, churn rate, renewal rate, or contract length for any Linkerbot customer relationship. | 低 | |
| CU017 | No reviewed source discloses customer-satisfaction survey data, Net Promoter Score, or any structured satisfaction metric for Linkerbot. | 低 | |
| CU018 | No reviewed source discloses repeat-purchase behavior (e.g., a named customer ordering a second or subsequent batch of hands) for any specific account. | 低 | |
| CU019 | The disclosed rise in monthly order volume from over 1,000 units (November 2025) toward a 10,000-unit/month target (mid-2026) is consistent with, but does not prove, some combination of new-customer growth and repeat/expansion orders from existing accounts, since no source breaks out new vs. repeat unit volume. | 中 | SU004, SU007 |
| CU020 | No reviewed source discloses customer-concentration data (e.g., top-customer or top-5-customer revenue share) for Linkerbot. | 低 | |
| CU021 | Linkerbot's disclosed go-to-market channel is direct sales plus international press/trade-show visibility rather than a disclosed reseller network, implying limited channel/partner dependence risk as currently understood, though this is inferred rather than directly confirmed. | 中 | SU003, SU009 |
| CU022 | Linkerbot's HKTDC international-sourcing storefront page and a Hannover Messe exhibitor page both returned not-found results as of the report date, indicating at least two international B2B trade-directory channels are not currently active or discoverable, a minor procurement-friction/channel-visibility signal for prospective international buyers. | 中 | SU024, SU025 |
| CU023 | China's National Development and Reform Commission publicly warned in 2026 that more than 150 humanoid-robot companies now exist domestically, an adverse sector-level signal relevant to assessing whether Linkerbot's customer base could face switching pressure from an increasingly crowded competitive field. | 中 | SU013 |
| CU024 | No reviewed source documents any customer complaint, blocked deployment, failed pilot, or contract cancellation involving Linkerbot specifically. | 低 | |
| CU025 | CB Insights maintains an analyst-tracked profile of Linkerbot, indicating independent commercial-data-provider coverage exists, though no customer-level detail from that profile was visible in the reviewed excerpt. | 中 | SU012 |
| CU026 | Linkerbot's hands are described in press coverage as capable of tasks including playing piano, threading needles, tightening screws, and assembling electronics, illustrating the breadth of use cases across research demonstration and industrial precision-assembly workflows. | 中 | SU016, SU006 |
| CU027 | No reviewed source provides a quantified outcome metric (e.g., cycle-time reduction, defect-rate improvement, cost savings) for any named customer's use of Linkerbot's hands; all customer proof found is logo-level (naming the customer) rather than outcome-level. | 中 | SU002, SU004, SU001 |
| CU028 | No reviewed source discloses a land-and-expand pattern (e.g., a customer starting with a pilot order and expanding to a larger production order) for any named Linkerbot account. | 低 | |
| CU029 | Given that Linkerbot's own headcount and R&D team plans reportedly include doubling its research team by the end of 2026 (per an independent press report), incremental customer-facing and support capacity may lag hardware production growth, an inferred but unconfirmed operational-capacity risk. | 中 | SU001 |
| CU030 | No reviewed source discloses whether Linkerbot offers different service/support tiers to different customer segments (e.g., research vs. industrial), so a segment-differentiated support model cannot currently be confirmed. | 低 | |
| CU031 | Sector-level coverage frames 2025-2026 as a period of rapid customer-base expansion across many funded Chinese dexterous-hand and humanoid-robot makers, a context in which Linkerbot's own named-customer roster (Samsung, Siemens, Stanford, MIT, Cambridge, University of Hong Kong) appears competitive in breadth but not verifiably larger than rivals' rosters. | 中 | SU015, SU010 |
| CU032 | No reviewed source discloses Linkerbot's total distinct customer count (logos), making it impossible to assess concentration risk as a percentage of total accounts rather than only in absolute/named terms. | 低 | |
| CU033 | The overall customer-evidence profile for Linkerbot combines credible, multiply-corroborated named-customer logos across industrial and research segments with a complete absence of retention, concentration, satisfaction, and outcome-level evidence, a pattern consistent with an early-scaling hardware company rather than a mature enterprise vendor. | 中 | SU001, SU002, SU004 |
| CU034 | An AOL republication of the same Reuters wire story includes a dateline photo caption placing Linkerbot's demonstration area at its Beijing office as of April 27, 2026, independently corroborating the company's Beijing headquarters location referenced elsewhere in customer- and adoption-facing coverage. | 中 | SU017 |
| CU035 | Humanoids Daily's coverage of China's broader 'robot fever' and repetitive-clone risk provides additional independent corroboration, beyond the SCMP NDRC report, that sector-wide overcapacity is a live adverse theme relevant to how durable Linkerbot's customer growth is likely to be. | 中 | SU018 |
| CR001 | US lawmakers introduced the bipartisan American Security Robotics Act and GUARD Act in March 2026, both aimed at banning US federal government procurement and use of humanoid and ground robots made by Chinese firms, a direct regulatory risk for any Chinese-made robotics component including Linkerbot's hands. | 高 | SR001, SR002, SR003 |
| CR002 | No reviewed source confirms whether the GUARD Act's or American Security Robotics Act's scope would extend to components (dexterous hands) embedded inside non-Chinese-branded robots, leaving the practical reach of this regulatory risk to Linkerbot's US-linked sales uncertain. | 低 | SR002 |
| CR003 | BIS's 2026 license-review process for AI chip exports to China and Macau introduces case-by-case review rather than blanket denial, but still constrains the advanced-compute components that could be used in higher-end Linkerbot hand electronics or associated AI/data infrastructure. | 中 | SR004 |
| CR004 | US export controls published in January 2025 on advanced computing items and AI model weights add compliance obligations relevant to any embodied-AI company (including Chinese dexterous-hand makers) that trains or exports AI models alongside hardware. | 中 | SR005 |
| CR005 | A law-firm analysis frames humanoid robots as entering 'a new era of risk' under existing product-liability, safety, and data-protection law, a general legal-exposure backdrop relevant to any dexterous-hand maker, Linkerbot included, even absent a company-specific case. | 中 | SR006 |
| CR006 | No reviewed source documents any litigation, enforcement action, regulatory investigation, or product recall involving Linkerbot specifically. | 低 | |
| CR007 | No reviewed source discloses any patent, trademark, or other intellectual-property filing or dispute involving Linkerbot. | 低 | |
| CR008 | China is actively drafting national humanoid-robot standards through a technical committee whose members include executives from Unitree and AgiBot but, per reviewed sources, not Linkerbot, raising a risk that Linkerbot could face standards shaped without its direct input. | 中 | SR013, SR012 |
| CR009 | No reviewed source confirms whether Linkerbot's hand products currently comply with, or have been certified against, ISO 10218 or ISO/TS 15066 industrial/collaborative robot safety standards. | 低 | SR009, SR010 |
| CR010 | US OSHA maintains general workplace robotics safety guidance that would apply to any US industrial deployment of Linkerbot-equipped robots, an additional compliance layer beyond ISO standards for US-based industrial customers. | 中 | SR011 |
| CR011 | The EU's AI Act regulatory framework adds a further compliance layer for any embodied-AI hardware, including dexterous hands, sold into European enterprise or public-sector buyers, though no reviewed source confirms Linkerbot has any current EU sales exposure. | 低 | SR031 |
| CR012 | China's National Development and Reform Commission publicly warned in 2026 that more than 150 humanoid-robot companies now exist domestically and cautioned against excessive product duplication and speculative bubbles, a sector-wide regulatory-tightening risk relevant to Linkerbot's continued favorable policy treatment. | 中 | SR014 |
| CR013 | No reviewed source discloses Linkerbot's manufacturing facility count, supplier redundancy, or component-sourcing detail, so single-supplier or single-facility concentration risk cannot currently be assessed directly, though the general absence of disclosure is itself an operational-risk signal. | 低 | |
| CR014 | No reviewed source discloses field reliability data (defect rate, return rate, mean time between failures) for any Linkerbot hand model in production use. | 低 | |
| CR015 | No reviewed source documents any product recall, safety incident, or outage involving Linkerbot hardware. | 低 | |
| CR016 | Scaling monthly production from approximately 5,000 toward a targeted 10,000 units, and separately toward a reported 50,000-100,000-unit full-year 2026 target, represents material operational execution risk given no disclosed capital-expenditure or capacity-expansion plan to support it. | 中 | SR018, SR020 |
| CR017 | Linkerbot's software/control stack depends on a specific, relatively current OS/framework combination (Ubuntu 22.04, ROS Humble, Python 3.10+), creating an operational dependency risk if that combination becomes outdated or if a security vulnerability is discovered in an underlying open-source dependency. | 中 | SR023, SR022 |
| CR018 | No reviewed source discloses cybersecurity practices, penetration-testing results, or vulnerability-disclosure policies for Linkerbot's SDK, firmware, or teleoperation data pipeline. | 低 | |
| CR019 | An independent industry tracker frames the broader Chinese dexterous-hand category as moving toward commoditization as component-manufacturing entrants (e.g., Zhaowei) and full-stack integrators (e.g., AgiBot) enter with standalone hand products, an operational/commercial risk of margin compression for Linkerbot specifically. | 中 | SR028, SR029 |
| CR020 | Linkerbot's capital providers include a mix of Chinese state-linked funds (Zhongguancun Science Park Fund), a state-linked financial institution (Bank of China Asset Management), and private/strategic investors (Ant Group, HongShan Group, Fosun Capital), creating dependency on continued institutional-investor confidence for further scale-up financing. | 中 | SR018 |
| CR021 | Listed company Dongfang Precision's disclosed ~4.36% direct equity stake in Linkerbot is a form of strategic/financial partner dependency whose continuity is subject to that public company's own disclosure and governance decisions, outside Linkerbot's direct control. | 中 | SR021 |
| CR022 | No reviewed source discloses Linkerbot's top-customer or top-5-customer revenue concentration, so key-customer dependency risk cannot currently be quantified beyond the qualitative observation that only a handful of named accounts (Samsung, Siemens, Stanford, MIT, Cambridge, University of Hong Kong) have been publicly identified. | 低 | |
| CR023 | Linkerbot's go-to-market appears to rely on direct sales rather than a disclosed reseller/distributor network, and two international B2B trade-directory listings (HKTDC, Hannover Messe) return not-found results, a minor channel-dependency and visibility gap for prospective international buyers. | 中 | SR024, SR025 |
| CR024 | China's semiconductor and advanced-computing supply chain is subject to escalating US export controls, creating a partner/supplier-dependency risk for any China-based robotics hardware maker, Linkerbot included, that relies on internationally sourced advanced components. | 中 | SR007, SR004, SR005 |
| CR025 | Public reporting names at most one or two individuals (CEO Alex Zhou in 2026 coverage; 'Founder and CTO Zhou Yong' in 2025 coverage) as Linkerbot's public face, and a later independent source ties both names to a single individual ('Alex Zhou Yong'), but no board, executive team, or succession structure is disclosed, creating material key-person and governance risk. | 中 | SR030, SR017, SR018 |
| CR026 | Plans reported by an independent source to double Linkerbot's research team by the end of 2026, alongside a 5-10x production-volume growth target for the same year, represent material organizational execution risk if hiring, onboarding, and quality-control capacity do not scale proportionally. | 中 | SR019 |
| CR027 | No reviewed source discloses Linkerbot's total headcount, so execution risk from rapid team growth cannot be benchmarked against a current baseline. | 低 | |
| CR028 | No reviewed source discloses Linkerbot's board composition, independent directors, or any formal governance oversight structure beyond the legal-representative registration on file with China's business registry. | 低 | SR026 |
| CR029 | No reviewed source discloses Linkerbot's cash on hand, monthly burn rate, or runway, so financial/model risk from continued capital-intensive scale-up cannot be quantified beyond the observation that the company has raised at least three disclosed rounds in under 18 months. | 低 | SR020, SR018 |
| CR030 | No reviewed source reconciles the ~RMB 1.5 billion February 2026 Series B round with the ~$3 billion May 2026 'Series B+' valuation, leaving capital-efficiency and dilution risk between the two events unclear. | 中 | SR020, SR018 |
| CR031 | No reviewed source discloses Linkerbot's gross margin, so the risk that current list pricing (e.g., the ~$920 O6 model) is below sustainable cost cannot be ruled out, particularly given a stated multi-year path toward sub-$70 unit pricing. | 低 | SR028 |
| CR032 | An adverse public dispute between Unitree and AgiBot over global humanoid-shipment leadership claims, reported by Humanoids Daily, illustrates that self-reported scale and market-share claims across this competitive set (including Linkerbot's own 80%+ share claim) carry credibility risk absent independent audit. | 中 | SR016 |
| CR033 | China's broader 'robot fever' narrative, including explicit central-government warnings about bubbles and repetitive product clones, is corroborated across at least two independent adverse sources (SCMP and Humanoids Daily), strengthening the sector-level risk signal beyond a single outlet's framing. | 高 | SR014, SR015 |
| CR034 | No reviewed source documents a specific mitigation, insurance policy, or contingency plan Linkerbot has adopted in response to any of the regulatory, operational, partner, or people risks identified in this chapter. | 低 | |
| CR035 | A monitorable thesis-break trigger for regulatory risk would be formal passage and enacted scope of the GUARD Act or American Security Robotics Act explicitly covering embedded components, which would materially reduce Linkerbot's addressable US-linked demand. | 中 | SR001, SR002, SR003 |
| CR036 | A monitorable thesis-break trigger for operational risk would be a confirmed product recall, safety incident, or major reliability failure reported for any Linkerbot hand model in production use. | 低 | SR028 |
| CR037 | A monitorable thesis-break trigger for financial/model risk would be a down round, valuation markdown, or public statement of missed production/revenue targets relative to the disclosed 2026 unit-shipment goals. | 中 | SR018, SR020 |
| CR038 | A monitorable thesis-break trigger for people/execution risk would be a confirmed departure of the company's sole publicly named executive (Alex Zhou / Zhou Yong) without a disclosed successor. | 中 | SR030, SR017 |
| CR039 | CB Insights maintains an analyst-tracked profile for Linkerbot, providing at least one independent ongoing monitoring source for future funding, leadership, or risk-relevant events. | 中 | SR027 |
| CR040 | No reviewed source discloses any insurance coverage (product liability, business interruption, key-person) held by Linkerbot. | 低 | |
| CR041 | The overall risk profile for Linkerbot combines high-visibility sector-level regulatory and overcapacity risk (well-corroborated by multiple independent sources) with a near-total absence of company-specific mitigation, certification, or contingency disclosure, a pattern consistent with an early-scaling private company rather than a risk-mature enterprise. | 中 | SR014, SR001, SR006 |
| CR042 | Independent US policy commentary (a Foundation for Defense of Democracies analysis and Tech in Asia coverage) corroborates that the US Senate was actively considering a federal procurement ban on Chinese robotics as of March 2026, reinforcing that this regulatory risk was live in more than one independent outlet rather than a single news report. | 中 | SR033, SR034 |
| CV001 | The investment thesis for Linkerbot rests on disclosed 80%+ share in a fast-growing, cost-collapsing high-DoF dexterous-hand category, corroborated production scale (10,000+ cumulative units shipped, targeting 10,000/month), and backing from major institutional investors at a ~$3 billion valuation. | 中 | SV001, SV022, SV016 |
| CV002 | The anti-thesis is that Linkerbot's market-share claim is self-defined and unaudited, its financial statements (revenue, margin, cash position) are entirely undisclosed, its customer base rests on a handful of logo-level named accounts with no outcome or retention data, and it operates in a sector China's own NDRC has flagged for overcapacity risk. | 中 | SV019, SV015 |
| CV003 | No reviewed source provides audited financial statements, an independent market-share audit, or named-customer outcome data for Linkerbot, so the investment thesis currently rests more on growth-narrative corroboration than on financial or commercial due-diligence-grade evidence. | 中 | SV015, SV016 |
| CV004 | Given the strength of the growth narrative but the near-total absence of financial and independently audited evidence, a 'research-more' recommendation with medium confidence, high risk rating, and 'stretched' valuation stance is more defensible than an outright buy or avoid call. | 中 | SV001, SV019, SV015 |
| CV005 | No reviewed source supports a specific target return, hold period, or exit multiple for Linkerbot, since no entry price, share class, or liquidation-preference terms are disclosed for any specific investor. | 低 | |
| CV006 | Linkerbot's current financing context is a 'Series B+' round that closed around late April/early May 2026 at approximately a $3 billion valuation, with a disclosed target of roughly $6 billion for the next round, though no source discloses the timing or instrument (private round vs. IPO) for that next round. | 高 | SV001, SV002, SV003 |
| CV007 | A February 2026 Series B round of roughly RMB 1.5 billion (about $210 million) is not reconciled in any reviewed source with the ~$3 billion Series B+ valuation reached roughly three months later, leaving entry-price discipline for any investor in that window difficult to assess independently. | 中 | SV004, SV001 |
| CV008 | No reviewed source discloses liquidation-preference terms, anti-dilution provisions, or any preference-stack detail for Linkerbot's investors, so preference overhang for a hypothetical new entrant cannot be assessed. | 低 | |
| CV009 | Listed company Dongfang Precision's disclosed ~4.36% direct equity stake in Linkerbot, confirmed via its own investor-relations disclosure, is the only quasi-independent, regulator-adjacent evidence point bearing on Linkerbot's actual valuation and cap table found in any reviewed source. | 中 | SV005 |
| CV010 | Public evidence (unit-shipment growth, named institutional investors, a technology-competition win) is directionally consistent with continued investor demand at a multi-billion-dollar valuation, but does not independently confirm that the specific $3 billion (or targeted $6 billion) figures are fair value rather than momentum-driven pricing. | 中 | SV026, SV001 |
| CV011 | In a bull case, Linkerbot sustains 80%+ share of a fast-growing high-DoF dexterous-hand category, reaches its 50,000-100,000-unit 2026 delivery target, and successfully raises at or above the disclosed $6 billion next-round target, supported by continued Chinese industrial-policy tailwinds. | 中 | SV001, SV004 |
| CV012 | In a base case, Linkerbot continues to grow shipments and revenue but at a more moderate pace than its stated targets, raises its next round at a valuation below the disclosed $6 billion target, and faces intensifying price competition from AgiBot, Zhaowei, and other standalone-hand entrants. | 中 | SV028, SV016 |
| CV013 | In a bear case, an independent audit or competitor dispute undermines confidence in Linkerbot's 80%+ share claim, China's NDRC overcapacity warnings translate into concrete policy tightening or subsidy withdrawal, and the company is forced into a down round or delayed fundraising as manufacturing scale-up falls short of its disclosed targets. | 中 | SV019, SV020 |
| CV014 | No reviewed source provides an explicit probability weighting across bull, base, and bear scenarios for Linkerbot; the scenario framing here is qualitative and evidence-anchored rather than a modeled probability distribution. | 低 | |
| CV015 | Unitree, a Chinese full-stack humanoid-robot maker with self-developed dexterous hands, reached approximately a $1.6 billion valuation after a mid-2025 Series C round and subsequently filed for a STAR Market IPO reported at up to $580 million in proceeds, providing an audited-filing-level comparable reference point for the broader China embodied-AI/robotics category. | 高 | SV006, SV007, SV008 |
| CV016 | Figure AI, a US-based full-stack humanoid-robot maker, raised more than $1 billion in a Series C round at a $39 billion post-money valuation, the highest disclosed valuation among reviewed comparables and a benchmark for best-case Western humanoid-robotics multiples. | 中 | SV009 |
| CV017 | Apptronik raised $520 million at a $5 billion valuation, and X Square Robot surpassed a $2.8 billion valuation across four consecutive rounds, both providing additional comparable reference points in the $2.8-5 billion range for embodied-AI/humanoid-robotics companies at a scale-up stage similar to Linkerbot's own. | 中 | SV010, SV011 |
| CV018 | AgiBot, a Chinese humanoid-robot maker and direct dexterous-hand competitor (OmniHand), was reported to have reached a valuation exceeding RMB 7 billion (roughly $1 billion) in an earlier funding round, positioning it below Linkerbot's disclosed ~$3 billion Series B+ valuation despite being a closer product-category comparable than the larger Western humanoid makers. | 低 | SV012 |
| CV019 | Against this comparable set, Linkerbot's disclosed ~$3 billion valuation sits above the closest direct product-category comparable (AgiBot, ~$1 billion) but below the largest full-stack humanoid-robot makers (Figure AI at $39 billion, Apptronik at $5 billion), suggesting the market is pricing Linkerbot closer to a broad humanoid-robotics platform than to a narrower hand-component supplier. | 中 | SV001, SV009, SV010, SV012 |
| CV020 | No reviewed source provides a disclosed revenue or EBITDA multiple for Linkerbot or for any of the reviewed comparables, so valuation benchmarking here is necessarily based on absolute round valuations rather than a standardized multiple. | 低 | |
| CV021 | Whole-market analyst estimates (MarketsandMarkets, Goldman Sachs Research) suggest the broader humanoid-robotics market could reach the tens of billions of dollars by the mid-2030s, a macro tailwind context for comparable valuations but not a company-specific input for Linkerbot's own price. | 中 | SV013, SV014 |
| CV022 | No reviewed source discloses any exit event (acquisition, IPO filing, or secondary sale) specific to Linkerbot; the closest evidence of exit readiness is comparable peer Unitree's own IPO filing, which demonstrates a viable exit path exists in the sector generally. | 中 | SV006, SV008 |
| CV023 | Given Linkerbot's disclosed scale (10,000+ units shipped, targeting far higher 2026 volumes) and institutional-investor base including large asset managers, an eventual IPO (plausibly on a Chinese exchange, mirroring Unitree) is a plausible but currently undisclosed and unconfirmed exit path. | 低 | SV006, SV001 |
| CV024 | No reviewed source discloses any lock-up terms, secondary-sale mechanism, or existing investor liquidity path for Linkerbot shareholders. | 低 | |
| CV025 | China's National Development and Reform Commission's 2026 warning about more than 150 domestic humanoid-robot companies and the risk of speculative bubbles is a direct adverse signal relevant to whether Linkerbot's rising valuation trajectory reflects durable fundamentals or sector-wide momentum. | 中 | SV019 |
| CV026 | A public dispute between Unitree and AgiBot over global humanoid-shipment leadership claims, reported by Humanoids Daily, illustrates that self-reported scale metrics across this competitive set — including Linkerbot's own unaudited 80%+ market-share claim — carry credibility risk that could affect valuation if contradicted by independent data. | 中 | SV020 |
| CV027 | US legislative moves (the GUARD Act, American Security Robotics Act) targeting Chinese-made robots and components are a further adverse valuation-risk signal, since they could constrain Linkerbot's addressable market in the US even as its China-domestic opportunity remains large. | 高 | SV021, SV030 |
| CV028 | No reviewed source discloses Linkerbot's cash position, burn rate, or gross margin, so a discounted-cash-flow or unit-economics-based valuation cross-check is not currently possible; the valuation stance in this chapter relies primarily on comparable-transaction and narrative-consistency analysis rather than fundamental modeling. | 低 | |
| CV029 | A monitorable thesis-break trigger would be an independent audit or customs/shipment-data analysis materially contradicting Linkerbot's claimed 80%+ high-DoF market share. | 中 | SV016, SV027 |
| CV030 | A monitorable thesis-break trigger would be a down round, materially delayed fundraising versus the disclosed $6 billion next-round target, or a public report of missed 2026 production targets. | 中 | SV001, SV004 |
| CV031 | A monitorable thesis-break trigger would be enacted US legislation explicitly banning components (not just complete robots) sourced from Chinese suppliers, which would materially shrink Linkerbot's addressable Western demand. | 中 | SV021, SV030 |
| CV032 | A monitorable thesis-break trigger would be departure of Linkerbot's sole publicly named executive (Alex Zhou / Zhou Yong) without a disclosed successor, given the concentrated key-person risk documented in the Risks chapter. | 中 | SV025, SV001 |
| CV033 | The single highest-value final diligence ask is an independent, third-party audit of Linkerbot's claimed market share and unit-shipment volumes, since nearly every other valuation input (comparable positioning, growth narrative, investor confidence) is downstream of that claim's credibility. | 中 | SV016, SV020 |
| CV034 | A second critical diligence ask is basic financial-statement access (revenue, gross margin, cash position, burn rate), without which no valuation method beyond comparable-transaction pricing can be applied with confidence. | 中 | SV015 |
| CV035 | A third diligence ask is reconciling the February 2026 Series B round size with the May 2026 Series B+ valuation to understand actual dilution and capital efficiency between the two events. | 中 | SV004, SV001 |
| CV036 | CB Insights maintains an analyst-tracked profile of Linkerbot, providing at least one independent ongoing data source that could be used to monitor future valuation-relevant events (new rounds, executive changes, or corporate actions). | 中 | SV015 |
| CV037 | Linkerbot's technology-competition win (Zhongguancun International Advanced Technology Competition, March 2026) and government-portal recognition provide qualitative, non-financial evidence of technical credibility that supports but does not independently confirm the disclosed valuation level. | 中 | SV026 |
| CV038 | Sector-level coverage frames 2025-2026 as a period of intense capital deployment across many Chinese and Western embodied-AI/humanoid-robotics companies, meaning Linkerbot's valuation trajectory should be read partly as a reflection of sector-wide capital availability rather than purely company-specific fundamentals. | 中 | SV029, SV017 |
| CV039 | No reviewed source models Linkerbot's valuation sensitivity to specific input changes (e.g., market-share haircut, production-target shortfall, or multiple compression), so any sensitivity analysis in this chapter is necessarily illustrative rather than derived from disclosed company data. | 低 | |
| CV040 | Overall, Linkerbot's valuation stance is best described as 'stretched relative to disclosed evidence': the growth narrative and investor roster are genuinely strong, but the near-total absence of audited financials, independent market-share verification, and reconciled funding-round data means the disclosed $3 billion (and targeted $6 billion) figures cannot currently be independently underwritten. | 中 | SV001, SV015, SV019 |
| CV041 | An independent market-research publisher (Research and Markets) and analyst company-profile trackers (Sacra for Figure AI, Tracxn for X Square Robot) provide additional independent corroboration for the comparable-set valuations used in this chapter, beyond the press releases and news wires already cited. | 中 | SV033, SV034, SV035 |
| CV042 | Crunchbase News, TechCrunch, and CB Insights each maintain 2026 unicorn-tracking coverage that frames robotics as one of several active categories minting new billion-dollar-valuation companies, corroborating that Linkerbot's rapid path to a multi-billion-dollar valuation sits within a broader, well-documented 2026 unicorn-formation trend rather than being an isolated outlier. | 中 | SV036, SV037, SV038 |