初创公司尽调
尽调报告 Cybersecurity / privacy-preserving computation Series B 2026-05-28

Zama

FHE 基础设施龙头,技术证明扎实,但商业化披露仍不完整

Zama 看起来是 FHE 机密区块链基础设施里最清晰的公开领导者,但当前估值跑在已披露商业化证据前面,因此更适合继续研究,而不是直接买入。

封面要素

总部 01
Paris, France [CO001]
Series B 轮估值 03
>1000 $M [CO014]
团队规模 04
96 [CO008]
主网上线 05
2025-12-31 [CE035]
Fairer 公开证据区间 06
550-840 $M [CV040]

公司概况

Zama 是一家总部位于巴黎的开源密码学公司,2019 年末注册,公开材料称由 Rand Hindi 和 Pascal Paillier 于 2020 年创立。 公司靠两条线把全同态加密(FHE)商业化:面向公链的保密层,以及涵盖 FHEVM、TFHE-rs、Concrete 和 Concrete ML 的更广泛开发者栈。 截至 2026-05-28 的运行日,公开证据显示团队 96 人、其中 37 名博士,并有一线加密与基础设施投资方支持,协议栈也已上线;但收入质量和客户耐久性的披露仍有限。

官网
zama.ai
成立时间
2020-01-01
创始人
Rand Hindi, Pascal Paillier
创立地点
Paris, France
总部
Paris, France
产品
Zama 销售分层 FHE 栈:Zama Confidential Blockchain Protocol 与 FHEVM 服务于既有链上的保密智能合约,TFHE-rs、Concrete 和 Concrete ML 则服务于加密计算与隐私保护机器学习。
客户
核心客群是区块链开发者、代币化资产发行方、钱包、托管与结算基础设施提供商,以及其他需要保密链上工作流的机构;次级方向包括隐私保护 AI 和加密数据应用。
商业模式
靠开源开发者采用、企业许可、协议 / 代币活动和伙伴主导的集成变现;公开证据确认多个变现面,但尚未确认已实现收入结构或利润率。
阶段
Series B
融资情况
2024 年 3 月完成 $73M Series A 轮,2025 年 6 月以超过 $1B 的估值完成 $57M Series B 轮;2026 年 1 月公开代币销售又带来 $44M 已披露资本。
[CO001, CO005, CO006, CO008, CO014, CE003, CE035]

执行摘要

主要优势

  • FHE 技术栈横跨 TFHE-rs、Concrete ML、FHEVM 和 Zama protocol,技术差异化清晰,开源开发者牵引也可见。
  • T-REX、Dfns、GSR、OpenZeppelin 和 J.P. Morgan Kinexys 等机构生态证据显示,公链上的合规保密确有需求。
  • 资本基础扎实,专业投资者加持,让 Zama 有现金跑道继续巩固机密基础设施品类领导位。
  • 团队研发底色很重(96 人、37 名 PhD),在难度很高的密码学领域支撑了技术可信度。

主要风险

  • 收入、毛利率、留存和客户集中度仍未披露,公开证据不足以承销商业化质量。
  • 面向受监管机构销售机密金融协议时,MiCA、Travel Rule、制裁和隐私义务都会带来真实执行风险。
  • 可见牵引很大一部分仍由合作伙伴、试点或公告驱动,能否转成持久生产使用尚未证明。
  • 估值标记从约 $550M 的代币销售定价到超过 $1B 的股权定价不等;如果经济性跟不上叙事,会形成压力。
  • 交易量增长后,运营方集中度、KMS 协调和性能扩展仍是重要技术与可靠性风险。

未决问题

  • 经验证的收入、费用消耗、企业合同价值和客户集中度数据。
  • 机密交易与协议运营的毛利率、补贴和服务成本证据。
  • 管理层或法律顾问材料,说明 MiCA、Travel Rule、制裁和隐私义务如何映射到运营流程。
  • 承销稀释和潜在压力所需的股权结构表、资金库、解锁,以及股权 / 代币互动细节。

目录

Chapter 01

01公司概况

1.1 身份与运营模式

在公司主体层面,Zama 的法律与商业身份异常清晰;到产品叙事层面,边界仍有些流动。 法律声明和法国登记记录显示,Zama 是一家注册办公室位于 8 rue du Sentier、总部在巴黎的 SAS;当前官网则不再把 Zama 包装成泛化的 FHE 研究实验室,而是强调以 Zama Protocol、$ZAMA 代币、开发者工具和企业区块链用例为中心的运营型协议公司。 这个公开姿态很关键:公司叙事已从宽泛的“开源密码学公司”转向协议与网络模式,后续章节也应把它作为当前运营基线。 同一批官方材料和代码仓库页面还显示,Zama 推的不是一条新链。它借助 FHE 库、协议服务、SDK 和运营商基础设施,在既有公链之上销售一层保密能力。 这一区别对尽调很重要,因为它决定买方采用路径、资本强度和竞争集合。 Zama 的开源足迹也不是装饰:组织页面、TFHE-rs、Concrete ML,尤其是 FHEVM,都显示出可量化的开发者关注度,说明公司的产品面不止一个代币或一条合作公告。[CO001, CO002, CO006, CO007, CO020, CO021]

KPI 快照表
指标数值 / 状态日期置信度缺口 / 限制
法律实体Zama SAS2026-05-28登记信息与法律声明中的实体名称一致
总部法国巴黎 75002,8 rue du Sentier2026-05-28公开办公地址,不证明员工集中在当地
创立 / 注册2020 年创立;2019-11-27 注册2024-03 至 2026-05创立叙事和法律注册时间并不以同一日期为准
团队规模96 人2026-05-28公司自称当前员工数
研究深度37 名博士 / 26 个国籍2026-05-28公司自称的人才指标
总融资额> $150M2025-06-25独立报道给出的 Series B 后总融资额
最新估值> $1B2025-06-25私募市场估值,不是公开市场出清价
早期商业信号声称已有数十家区块链和 AI 授权客户2025-06-25没有公开客户名单或合同金额明细
GitHub 数据组织关注者 3.7k;fhevm 星标 25.3k2026-05-28开发者信号,不是直接收入证据
收入 / 年化收入2026-05-28留存来源未披露公开审计收入或 ARR
准确客户数2026-05-28公开信息只有“数十家公司”和伙伴渠道表述
协议推出公开测试网已上线;Ethereum 主网 / 代币状态在不同来源中表述不一2025-12 至 2026-05官方、伙伴和交易所摘要使用的时间表述不同

缺少支持的商业覆盖指标保留为 null 或附限制说明;开发者指标作为牵引力代理纳入,不替代财务指标。

[CO001, CO002, CO008, CO014, CO017, CO019]
FO002: 公司快照逻辑

Zama 的法律身份、开源技术栈、融资基础、运营方网络和机构合作如何串起当前公司故事。

本图为定性图,表达业务逻辑,不代表持股比例或资金流。

[CO020, CO021, CO027, CO030, CO033, CO036]
FO003: 快照 KPI

截至运行日期,Zama 有证据支撑的经营和生态指标。

开发者信号和公司声称的商业指标只是采用度代理,不是经审计财务表现。

[CO008, CO017, CO019, CO023, CO024, CO025]

1.2 创始人、领导层与治理

创始人质量是 Zama 最强的公开可见资产之一。 Rand Hindi 仍是公司在法律披露、融资公告和产品传播中的主导外部发声者;Pascal Paillier 则为这套栈压住科学可信度。 合作方和媒体来源强化了同一条叙事:Zama 的可信度很大程度上靠这对创始人支撑。Hindi 把公司的隐私逻辑翻译成市场故事,Paillier 则提供密码学权威,让“FHE 已从理论走向可用基础设施”这句话站得住。 公开团队快照也显示,Zama 的研究属性异常重:96 名员工中有 37 名博士;Jeremy Bradley 带出可见的运营层,Benoît Chevallier-Mames 则体现产品工程深度。 但机构投资者通常想看的结构性披露仍偏薄。公开证据足以识别创始人和少数运营者,却不足以画出完整董事会、投票结构或当前股权结构表控制权。 结果是:这家公司看起来由创始人主导、人才密度很高,但治理透明度仍不足。因此,在进入代币或协议经济之前,关键人依赖和信息不对称已经是真实尽调议题。[CO003, CO004, CO005, CO008, CO009, CO010]

领导层与创始人表
人物职责有证据支持的背景职能覆盖关键人物依赖
Rand Hindi联合创始人、CEO融资、产品叙事和隐私论点的公开代言人;据 TechCrunch,曾创办 Snips。战略、融资、外部定位
Pascal Paillier联合创始人、密码学负责人官方和伙伴材料均列为创始人;定位为核心 FHE 权威。密码学研发可信度
Jeremy BradleyCOO / 总经理法律声明和 INPI 记录将其列为运营高管。运营与公开发布治理
Benoît Chevallier-Mames(人物)产品工程副总裁AWS 案例研究称其为基础设施和产品工程负责人。产品工程与部署

覆盖不完整,因为公开来源没有披露完整董事会、委员会结构或完整高管梯队。

[CO003, CO004, CO005, CO009, CO026]

1.3 资本基础与合作节奏

Zama 的融资曲线陡到本身就值得关注。 官方与独立来源口径一致:2024 年 3 月,Multicoin Capital 和 Protocol Labs 领投 $73 million Series A 轮;2025 年 6 月,Pantera Capital 和 Blockchange Ventures 领投 $57 million Series B 轮,估值超过 $1 billion。 独立报道把这次跃升视为首个 FHE 独角兽里程碑,管理层则把后一轮描述为战略性辛迪加搭建,而不只是财务融资。 这些说法有可信度,因为投资人名单与公司的市场进入目标相匹配:Zama 想让公链保密能力成为新的机构级原语,因此招来的投资人与合作方本就扎在区块链金融里。 Series B 之后的合作节奏支撑了这一逻辑。公开公告显示,T-REX、GSR、OpenZeppelin、JPMorgan 的 Kinexys 概念验证和 Dfns 以不同角色进入 Zama 生态——分销、开发者工具、市场结构和运营商基础设施。 重要的是,它们并不等价。有些是概念验证,有些是分销或工具关系,有些是合作伙伴对网络设计的验证。 但合在一起看,Zama 用资本不只是延长现金跑道,也在围绕保密公链工作流搭建一条可信的机构叙事。[CO011, CO012, CO013, CO014, CO015, CO016]

利益相关方 / 投资者图谱
利益相关方角色控制权 / 经济重要性证据尽调问题
Multicoin CapitalSeries A 领投方2024 年帮助奠定区块链金融投资者基础。Series A 公告获取持股比例、按比例跟投权和董事会权利
Protocol LabsSeries A 领投方撑起密码学 / 基础设施可信度和早期融资深度。Series A 公告确认战略产品重叠度与当前参与度
Blockchange VenturesSeries A 参投方、Series B 共同领投方连续参投显示其是最有信念的支持方之一。Series A + Series B 报道确认后续持股和治理权利
Pantera CapitalSeries B 共同领投方带来大型加密分发能力和机构信号,支撑协议商业化。Series B 公告厘清代币相关权利与股权权利
战略创始人投资团天使 / 生态投资者Juan Benet、Gavin Wood、Anatoly Yakovenko 等带来的网络效应大于纯资本。Series A 公告判断战略资源是否转化为活跃合作
Dfns / 运营商伙伴网络运营商和分发渠道为 MPC/KMS 提供运营可信度,也可能触达企业客户。Dfns 运营商文章与公司保密报告区分运营商经济性与真实终端客户需求

覆盖不完整,因为公开来源识别了投资者名称和伙伴角色,但没有披露持股比例、董事会席位或代币权利条款。

[CO011, CO012, CO014, CO015, CO018, CO027]
FO001: 公司里程碑时间线

时间线展示 Zama 从法国注册和 2020 年创立,到融资、协议上线、机构合作以及不利代币市场信号的路径。

由于保留来源没有锚定具体日期,图中采用月份精度。

[CO002, CO011, CO014, CO017, CO028, CO030]

1.4 规模信号与关键警报

Zama 当前画像里最强的正面信号,是多个独立领域指向同一个模式:开源参与度很高,一线加密与基础设施投资方在场,合作伙伴贴近受监管金融工作流。 最强的负面信号,是市场信心仍跑在已验证运营披露之前。 Zama 称其向数十家公司许可技术,合作伙伴材料也指向机构用例;但公开证据仍撑不起收入、精确客户数或董事会层面治理的硬指标。 同时,多家独立来源认为底层 FHE 市场仍处早期,可用性能仍离不开专用硬件进步,公开代币价格波动也足以暴露估值风险。 TechCrunch 质疑同态加密是否已达到大众市场可扩展性;MEXC 和 CoinGabbar 则把公开代币安排描述为估值激进、面临供应悬顶。 这些都不是致命警报,但会改变尽调姿态。 Zama 目前更像一家资本充足、技术差异化明确、确有生态拉力的协议公司,而不是一家已完全去风险、披露纪律成熟的金融基础设施平台。[CO019, CO028, CO029, CO036, CO037, CO038]

里程碑表
日期事件类型金额 / 状态参与方影响
2019-11-27法国注册信息出现在 INPI 记录中创立SIREN 879243319Zama SAS法律实体早于公开的 2020 年创立叙事
2020创立叙事成为公开基准创立公司成立Rand Hindi;Pascal Paillier确立伙伴和媒体反复沿用的创始时期起点
2024-03-07Series A 完成融资$73MMulticoin;Protocol Labs;生态天使为商业化推进和招聘提供资金
2024-03-07TechCrunch 报道已签约合同价值超过 $50M、团队 75 人规模早期商业化信号TechCrunch 访谈表明在协议叙事完全成熟前已有需求
2025-06-25发布 Series B 和 Confidential Blockchain Protocol融资$57M,估值 $1B+Pantera;Blockchange让 Zama 成为私募市场独角兽,并把叙事转向协议商业化
2025-06-25凸显公开测试网 / 主网推进序列产品协议开始推出Zama 管理层标志着公司从库公司转向实时网络基础设施
2026-03-26T-REX Ledger 保密合作宣布合作$32B ERC-3643 资产;2027 年 6 月前目标 $100BT-REX;Apex 生态;Zama验证机构级代币化用例
2026-03Dfns 加入运营商网络合作Genesis MPC 节点角色Dfns;Figment;InfStones;其他说明运营商网络建设已超出公司自营
2026竞争性隐私模型仍活跃治理ZK 和许可型技术栈也在争取机构客户Cointelegraph 访谈证实市场仍在争夺中
2026-02 起公开代币上市后走弱负面后续评论称跌幅接近 50%CoinGabbar;MEXC;Phemex给公司故事带来估值和代币经济包袱

时间线有意精选,只覆盖会影响后续章节假设的创立、融资、产品、合作和负面事件。

[CO002, CO011, CO014, CO017, CO027, CO030]

1.5 图表

Chapter 02

02市场分析

2.1 市场边界与现状替代方案

Zama 的市场只有窄定义才说得通。 公司不是泛泛参与“隐私”,甚至也不是覆盖整个机密计算市场。它销售的是可编程保密能力,服务于仍必须存在于公链上的资产和逻辑。 这个边界很重要,因为它排除了大量相邻支出:如果组织更在意控制权、结算终局性或法律可执行性,而不是共享公链可组合性,仍可选择私链、重零知识设计、可信执行环境或许可网络架构。 保留来源从不同角度反复描述同一个问题:公开账本会暴露敏感余额、投资人头寸、交易逻辑和身份关联数据;开放性有利于验证,却挡住许多机构或受监管工作负载。 因此,现状替代方案不是“什么都不做”,而是继续留在私有基础设施上,或采用范围更窄的隐私工具。 Zama 的相关性主张是:买方既能保留公链可组合性,又能隐藏数据。若这一主张成立,市场切入点就有意义;若不能,公司可能被夹在许可制既有厂商与复杂度更低、足以满足买方的替代隐私架构之间。[CM001, CM002, CM003, CM004, CM026, CM029]

市场定义表
细分市场 / 类别纳入的支出或价值排除的支出买方 / 付款方与 Zama 的相关性
公链保密支付公链上的稳定币支付流和支付基础设施通用卡处理、银行核心系统和仅限私有账本的支付技术栈钱包、支付提供商、发行方
公链上的代币化 RWA有保密需求的国债、商品、基金、证券和代币化资产基础设施仅限私链的证券化或链下基金管理资产管理人、发行方、代币化平台
保密 DeFi 与链上交易AMM、借贷、兑换、密封竞价拍卖和私密订单流不重视隐私的透明 DeFi,以及中心化交易所订单簿协议、做市商、资金管理团队
身份与治理工作流KYC 关联凭证、链上身份检查、保密投票常规链下 IAM 和非区块链投票系统应用、DAO、受监管运营方
更广义的机密计算与区块链保密相邻的企业隐私基础设施预算非区块链机密计算工作负载云和数据安全买家低 / 仅相邻

纳入市场刻意窄于全部机密计算或全部代币化;相关性评分反映其与 Zama 当前产品姿态的匹配度。

[CM001, CM002, CM003, CM004, CM005, CM025]
FM004: 采用漏斗或价值链地图

从开发者或发行方兴趣走向保密公链生产部署的可能路径。

该流程是概念性的,捕捉保留证据所暗示的商业化顺序,而不是实测转化漏斗。

[CM022, CM024, CM030, CM039, CM040]

2.2 规模测算视角:支付、代币化与机构金融

测算 Zama 市场,正确方法是叠加多个视角,而不是锚定一个膨胀的标题数字。 支付视角已经很大:litepaper 把稳定币视为领先的区块链用例,年交易量以数万亿美元计;Forbes 引用 a16z 研究称,仅 2026 年 Q1 稳定币支付量就达到 $4.5 trillion。 代币化视角今天更小,但与保密需求更明显对齐。Crowdfund Insider 汇总的 CoinGecko 数据显示,RWA 在 2025 年初以来增长两倍后,到 2026 年 Q1 达到 $19.3 billion;美国国债仍占主导,但大宗商品、股票和 ETF 增长很快。 机构基础设施又给出一条测算线索:T-REX 称 ERC-3643 已保护 $32 billion 资产,Apex 支持的基础设施目标是在 2027 年中达到 $100 billion。 更长期的总可用市场(TAM)口径差异巨大。McKinsey 的基准情景到 2030 年仍低于 $2 trillion;仅 BCG 支持的代币化基金材料就指向 $600 billion AUM,更宽的预测汇总远高于此。 正确结论不是所有万亿美元级估算都错,而是 Zama 最终 TAM 可以很大,但可信的近期可服务市场(SAM)要窄得多,并且绑定具体产品化工作流。[CM005, CM006, CM007, CM008, CM009, CM010]

TAM/SAM/SOM 或规模测算视角表
发布方年份地域 / 范围价值CAGR / 增长方法置信度限制
Forbes 引用 a16z2026全球稳定币支付流$4.5T(2026 年 Q1)N/A以季度支付追踪视角衡量活跃稳定币流支付流不等同于保密服务提供商的收入池
Crowdfund Insider 汇总 CoinGecko2026不含稳定币的全球代币化 RWA$19.3B(截至 2026 年 Q1)较 2025 年初 +256.7%主要代币化 RWA 类别的市值快照当前市场规模,不是长期预测
Chainwire / T-REX2026ERC-3643 生态和 Apex 目标当前 $32B / 2027 年 6 月前目标 $100BN/A机构代币化平台指标平台特定指标,不代表整体市场
BCG × Aptos × Invesco2024代币化基金2030 年前 >$600BN/A基于代币化基金达到共同基金与 ETF AUM 1% 的 AUM 预测仅针对基金,不覆盖全部资产
McKinsey 经 Ledger Insights 转述2024传统资产,不含稳定币2030 年前 $1T-$4T;基准情景低于 $2TN/A保守的机构代币化情景对 McKinsey 的二手报道,不是直接来源文本
AssetTokenization.com 综合2025跨发布方代币化预测2030/2034 年前 $1.9T 至 $30T+因来源而异主要机构预测汇编方法混杂,范围更宽

数值混合了流量、当前市值、平台目标和长期 AUM / 市场规模预测;只能作为叠加的规模测算视角比较,不能相加成一个总量。

[CM006, CM009, CM014, CM015, CM016, CM017]
FM001: 市场规模测算视角

受证据约束的金字塔,把广义代币化预测收窄为更现实的短期保密公链切入口。

所有数值均为 USD 十亿美元,代表不同市场视角,不能相加;底层是短期代理,而非 Zama 收入。

[CM014, CM015, CM016, CM038]
FM002: 市场估计区间

跨当前 RWA 规模和长期代币化预测的低 / 基准 / 高区间,以 USD 十亿美元表示。

所有数值均为 USD 十亿美元。最后一行混合了不同发布方的多年期乐观估计,只作为情景区间展示。

[CM009, CM016, CM017]

2.3 买方图谱、用户图谱与采用路径

买方图谱比一次简单企业软件销售复杂得多。 在发行方主导的工作流里,付费预算所有者可能是需要合规结算和保密投资人处理的资产管理人、基金发起方或代币化平台。 在运营商或钱包工作流里,经济买方也可能是托管或钱包基础设施提供商,它们必须为企业客户支持保密余额、KMS 操作或阈值解密服务。 开发者构成第三类:他们通常先是用户,之后才成为直接买方;只要保密能力能叠进 Solidity 和既有链,而不必从零学习新体系,他们就会采用工具。 这也是 OpenZeppelin 和 Dfns 关系重要的原因。两者在开发者体验和运营信任两个关键点降低采用摩擦。 因此,可能的采用路径不是一夜横向铺开,而是一条一条生态推进:先是工具包和试点,再是运营商参与,再到发行方和交易工作流,最后才是更广泛的默认采用。 这条路径与保留证据一致,但也意味着商业化会不均匀。与 Kinexys 的概念验证或与 T-REX 的合作公告能提高可信度,却不能自动证明所有买方细分都能重复转化预算。[CM019, CM020, CM021, CM022, CM023, CM024]

细分市场 / 买方图谱
细分市场买方用户付款方工作流预算负责人采用触发因素
资产发行方 / 代币化平台基金管理人或发行方运营、合规、过户代理发行方或平台保密发行、结算、投资者隐私业务线 GM / COO需要使用公共结算轨道,同时不暴露头寸
钱包 / 托管基础设施托管或钱包服务商运营、安全、企业客户基础设施提供商加密余额、KMS、策略驱动解密CTO / 产品负责人企业客户需要隐私功能
开发者工具和应用协议或应用构建方智能合约开发者协议金库或初创公司Solidity 和现有链上的保密合约创始人 / CTO借助库和标准降低集成摩擦
机构交易工作流交易场所或市场参与者交易、风控、合规团队交易场所或做市商私密执行、密封竞价、保密头寸市场 / 产品负责人需要限制信息泄露
身份 / 合规工作流受监管应用运营方合规官、终端用户运营方选择性披露和可验证凭证首席合规官 / 产品负责人需要兼顾隐私和可审计性
运营方网络基础设施运营方节点 / enclave 运营方运营方自身或协议激励参与 MPC KMS 和协处理器基础设施总经理有吸引力的运营方经济性和生态增长

早期,买方、用户和付款方往往合并为同一个主体,但发行方和运营方工作流会把决策权和 经济激励拆开。

[CM019, CM020, CM021, CM022, CM023, CM024]
FM003: 买方 / 细分市场地图

定性地图,说明每类买方中保密性最能解决哪些痛点。

这是定性矩阵;色调概括证据强度,而不是数值份额。

[CM019, CM020, CM021, CM022, CM024, CM039]

2.4 增长驱动、约束与市场风险

Zama 的需求侧逻辑很直接:公链金融规模已经很大,代币化正在扩展,更多机构工作流想要隐私加互操作性,而不是用隔离换隐私。 保留来源显示,合规、选择性披露和抗抢跑不是附带收益,而是买方愿意付费的核心原因。 但约束同样重要。技术仍要靠专用硬件进步、运营商网络信任和性能提升,才能补上“演示很有吸引力”与“默认市场标准”之间的差距。 竞争性隐私模型仍然可行。Zama 自身代币又引入第二个市场变量:如果估值、解锁和投机压过使用增长,就可能反噬企业可信度。 这里保留的反向证据并不致命,但足以否定无摩擦采用故事。 最现实的看法是:Zama 处在一个增长市场中,结构性需求驱动很强;但公司仍要证明基于 FHE 的保密能力可以经济化扩展、顺畅集成,并在技术周期和代币市场周期中持续守住信心。[CM025, CM027, CM028, CM029, CM030, CM031]

增长驱动因素与约束表
驱动因素 / 约束方向时间含义尽调问题
稳定币支付规模驱动当前公链支付流量够大,保密性开始具备经济意义这部分流量里,哪一块真正需要隐私控制?
RWA 代币化增长驱动2026-2030机构资产上链,带来明确的保密需求代币化有多少会留在公链轨道,多少转向私有轨道?
选择性披露与合规驱动当前有审计路径时,隐私功能更容易被采用哪些司法辖区接受运营方 / 解密模型?
借助 Solidity 和标准化工具驱动当前开发摩擦更低,可能比定制密码学栈更快扩大采用从开发者到付费部署的真实转化率是多少?
对专用硬件的依赖约束当前-2027性能经济性仍取决于 GPU 和未来加速器生产规模下,每笔保密交易成本是多少?
竞争性隐私架构约束当前ZK、TEE、许可网络等模型争夺同一批买方预算Zama 相比每个替代方案,在哪里有决定性胜出?
运营方网络复杂度约束当前采用前提是信任 KMS、协处理器和运营可靠性有多少运营方?正常运行时间和罚没保护机制是什么?
代币悬压与波动约束当前投机性代币重定价可能削弱企业信心相比解锁供应,使用驱动需求有多少?
机构验证点驱动近期T-REX 和 Kinexys 让市场叙事更可信试点是否转成经常性生产量?
预测不确定性约束持续预测区间从低于 $2T 到高于 $30T,投资人容易围绕醒目的 TAM 数字过拟合即使在保守情景下,哪些自下而上的 SAM 假设仍站得住?

行内同时放入结构性驱动和执行约束,因为商业结果取决于前者多快跑赢后者。

[CM006, CM009, CM015, CM025, CM027, CM029]

2.5 图表

Chapter 03

03竞争格局

3.1 格局与替代边界

不能只拿一个狭窄同业集合来承销 Zama。 直接同业是一小群 FHE 原生保密层公司,尤其是 Fhenix 和 Inco,因为它们同样承诺为既有区块链用户提供保密智能合约。 替代集合更宽,也更危险。Secret Network 和 Oasis Sapphire 提供已经推向市场、信任假设不同的保密执行环境;Aztec 和 Aleo 则借助新的 zk rollup 或专用私密应用栈,进攻相邻隐私工作负载。 Partisia 和 Arcium 又通过基于 MPC 的机密计算扩大竞争场。 现状替代方案往往根本不是另一个加密协议,而是 AWS、Azure 或 Google Cloud 的企业 enclave,让买方把敏感逻辑留在链下或内部构建工作流里。 更宽的边界很重要,因为买方不需要专门相信 FHE 就能解决同一个任务。若 TEE、zk-first 设计或云 enclave 栈看起来更成熟、更便宜、更容易向内部解释,采购团队仍可选择它们。[CP001, CP011, CP015, CP017, CP020, CP024]

竞争者画像表
竞争者类别信任模型执行环境目标买方 / 用例规模或成熟度证据相比 Zama 的关键限制
Zama参考公司FHE + 链下协处理器 + 阈值 KMS现有 L1/L2 链上的保密层希望在现有链上获得密码学保密性的机构和开发者Kinexys 概念验证;Dfns 分发;96 人团队真实市场规模仍早于成熟云服务或主网替代方案
Fhenix直接 FHE 同行纯 FHE,营销口径是不靠 TEE / ZK 变通带标准 Solidity 钩子的 Ethereum 隐私层想要隐私、但不想换新 VM 或 rollup 的 Ethereum 团队官网和文档强调 FHE 原生智能合约公开成熟度和机构验证少于 Zama 借 Dfns 与 Kinexys 声称的水平
Inco直接隐私层同行基于 TEE 的私有计算兼容钱包和 EVM/SVM 工具的隐私层优先看兼容性和广泛钱包 / 工具支持的开发者官网列出支持 MetaMask、Coinbase Wallet、Phantom、Hardhat、Foundry 和 Anchor硬件信任假设弱于 Zama 的纯 FHE 叙事
Secret Network已成型替代方案Intel SGX 加密钥管理独立的隐私优先网络想要上线可用私有智能合约、且看重更长主网历史的开发者文档称私有智能合约自 2020 年 9 月起上线需要押注独立网络和 TEE 信任模型
Oasis Sapphire已成型替代方案基于 TEE 的保密 EVM保密 EVM ParaTime优先看低费用和快速终局性的 EVM 团队官方文档宣传费用低 99%+、终局性 6 秒依赖 TEE 假设,而非纯密码学保密性
Azteczk 优先替代方案zk rollup,设备侧证明,加密 UTXO带新 VM 的隐私优先 L2相比迁移简单性,更看重反 TEE 隐私保证的开发者官方文档和网站宣传隐私优先的 Ethereum rollup 定位不兼容 EVM,因此迁移成本明显更高
Aleo专用私有应用平台私有应用栈,带自有程序和节点 / 质押流程专用私有应用网络愿意为隐私采用新编程和节点环境的团队文档覆盖程序、钱包、质押和漏洞赏金新技术栈和不同开发者路径提高切换成本
AWS / Azure / Google 保密计算维持现状的内部自建替代方案云 enclave / TEE 基础设施通用保密计算服务在现有云资产内解决同一保密问题的企业成熟采购和运营方熟悉度不是 Zama 所追求的区块链原生和可组合形态

行内概括保留下来的高信号替代方案,而不是市场上每个隐私项目。规模和成熟度证据仅限已抓取来源中 公开可见的信息。

[CP001, CP009, CP010, CP011, CP015, CP017]
FP001: 按信任最小化和集成摩擦划分的竞争定位

基于主要保留竞争对手提供的加密信任最小化程度和迁移摩擦,做出的序数地图。

轴值是分析师基于公开产品和文档证据给出的序数判断,不是基准测试得分。

[CP001, CP011, CP015, CP017, CP020, CP024]

3.2 能力、信任模型与开发者迁移

买方既看重不依赖可信硬件的隐私,又不想迁移到专用私链时,Zama 的技术故事最强。 Zama 自身材料和 Figment 分析都把产品描述为既有 L1 和 L2 之上的一层:用链下 FHE 协处理器加阈值密钥管理,而不是新执行环境。 Fhenix 是理念上最接近的对手,因为它同样为 Ethereum 营销纯 FHE 隐私,并称开发者可以留在标准 Solidity 中。 Inco、Secret、Oasis 和云 enclave 厂商则从不同角度竞争:它们承诺兼容性和更容易集成,但依赖 TEE 或 enclave 假设。 Aztec 从相反方向竞争,完全拒绝 TEE,改用隐私优先的 zk rollup;但这种方案会带来新 VM 和更高迁移成本。 Aleo 在架构上离 Zama 更远,因为它要求开发者进入自己的程序、节点和质押世界。 因此,能力问题不只是“有没有隐私”,而是买方愿意接受哪一种信任与迁移取舍。[CP002, CP004, CP012, CP013, CP014, CP016]

特性 / 能力矩阵
购买标准ZamaFhenixIncoSecret / OasisAztec / Aleo云 enclave
在现有链上运行部分部分不适用,链下
核心设计避开可信硬件Aztec 是 / Aleo 部分
主流 Solidity 风格开发路径Oasis 是 / Secret 部分通用 API
成熟采购熟悉度
更长真实市场成熟度早期早期早期更高不一
机构金融定位低至中
公开定价清晰度

单元格是有证据支撑的标签,不是基准评分。「部分」表示来源集显示有实质重合,但与 Zama 定位并不完全对等。

[CP002, CP012, CP013, CP016, CP018, CP023]
FP002: 功能广度 / 能力地图

热力图展示 Zama 在哪里领先、哪些替代方案更容易采购,以及公开证据哪里仍有限。

标签概括保留证据,不是经审计的基准测试结果。

[CP012, CP013, CP016, CP018, CP023, CP024]

3.3 定价可见度、分销能力与机构触达

这个市场整体公开定价都偏弱,这本身就是竞争事实。 Zama 称部署免费,但没有为保密应用公布清晰的企业价格表;许多同业协议同样不透明。 公开集合里的少数具体数字来自替代方案或基础设施厂商。Oasis Sapphire 公布了强相对性能和费用主张;AWS、Azure 和 Google 则把保密能力作为既有基础设施目录的一部分销售,而不是作为独立隐私产品。 Zama 很难匹配这种捆绑式分销能力。 同时,Zama 进入市场并非没有现场证据。Dfns 集成把保密代币标准暴露给 400 多家企业客户,Kinexys 概念验证则显示至少一个大型金融基础设施沙盒测试过这套栈。 这些是有意义的市场进入信号,但相较超大规模云基础设施或上线历史更长、市场运行更成熟的主网替代方案,它们仍更早、更窄。[CP003, CP006, CP009, CP010, CP016, CP021]

定价 / 打包对比
提供方公开定价依据公开可见内容仍未知内容竞争含义
Zama无公开企业价目表部署被描述为免费;产品运行在现有链上面向机构应用和运营方的实际成交价格低门槛有助采用,但企业经济性不透明,削弱买方建模
Fhenix未保留公开标价隐私即服务定位,以及无需新 VM 的集成叙事合同金额、计价单位和毛利率结构功能直接重叠,但缺少清晰公开价格锚
Inco未保留公开标价横跨钱包和开发框架的广泛兼容性叙事商业定价和生产支持条款即使经济性未披露,工具覆盖面也可能赢下试点
Secret / Oasis相对主张,而非可比标价Oasis 发布费用低 99%+、终局性 6 秒的主张;Secret 强调默认私有执行企业净成交价格和支持经济性即使缺少同口径定价,成熟替代方案看起来仍更容易采用
Aztec / Aleo未保留公开可比价目表安全和隐私姿态比商业打包更可见迁移经济性和生产支持定价对部分买方,信任模型差异可能比标价更重要
AWS / Azure / Google基础设施用量目录保密计算在更大的云资产组合里销售实现等效区块链保密性的特定工作负载成本捆绑采购在买方便利性上可能胜过协议原生隐私

这是一张打包表,不是已实现收入表。多数隐私协议公开材料在定价上仍远弱于架构或开发者体验。

[CP003, CP013, CP021, CP022, CP031, CP032]

3.4 护城河耐久性与反向竞争逻辑

真正耐久的问题是:Zama 的纯 FHE 保密能力会成为买方必备要求,还是仅仅成为几种可接受架构中的一种。 正向逻辑很清楚。Zama 结合了研究密集型团队、跨链姿态,以及一种避开可信硬件、同时保留熟悉开发者界面的隐私模型。 如果市场认定机构级保密能力需要密码学保证,而不是基于 enclave 的保证,Zama 的位置就异常有利。 反向逻辑在公开证据中同样清晰。ChainSafe 和 BlockEden 都把 2026 年市场描述为多种架构之间仍在进行的选择,而不是赢家通吃式转向 FHE。 Aztec 从信任角度明确攻击基于 TEE 的系统;AWS、Azure 和 Google 则能凭借既有采购路径和运营熟悉度赢下内部构建或受监管买方。 实务中,Zama 的护城河因此是有条件的:当买方想在既有链上获得强密码学隐私时,护城河会增强;当足够好的隐私、更容易采购或更成熟的生态比架构纯度更重要时,护城河会变弱。[CP007, CP008, CP026, CP037, CP038, CP039]

护城河耐久性 / 竞争风险登记表
护城河主张威胁严重性证据重要性尽调问题
现有链上的纯 FHE 保密性Fhenix 在 Ethereum 上给出类似承诺Fhenix 也用标准 Solidity 钩子宣传纯 FHE 隐私直接同行把差异缩窄到执行质量和生态牵引力在匹配试点中测试开发者对 Zama 与 Fhenix 的偏好
不要求可信硬件Aztec 认为 TEE 不可接受,zk 系统可以占住信任最小化叙事Aztec 明确攻击 TEE 和后门当 zk 迁移可接受时,反 TEE 买方并不会自动归 Zama访谈买方:zk 隐私是否值得迁移到新 VM
现有链部署云保密计算可在现有采购渠道内、以链下方式完成同一任务AWS、Azure 和 Google 都把保密计算作为通用基础设施销售维持现状的内部自建可能比区块链原生隐私更容易获批建模预算负责人有多常来自云基础设施,而不是协议创新
机构市场路径更宽泛的替代方案有更长生产历史或更强采购入口Secret 和 Oasis 有真实市场历史;超大规模云厂商已有现成合同企业采购里,分发可能比架构纯度更重要索取 Dfns 和具名机构试点的转化数据
研究型团队和 FHE 路线图硬件和性能路线图耗时可能超过替代方案采用曲线Zama 仍指向 GPU、FPGA 和 ASIC 里程碑,用来缩小吞吐差距如果买方今天接受足够好的保密性,未来硬件胜利可能来得太晚获取当前单笔交易成本,以及硬件里程碑时间线的信心
跨架构市场教育到 2026 年,尚无单一隐私架构显出主导地位ChainSafe 和 BlockEden 都把市场描述为 FHE、ZK、TEE 之间的取舍市场碎片化支持多栈并用,削弱硬锁定测试客户计划单栈标准化,还是分层共存

严重性来自分析师判断,反映投资判断风险,不是供应商报告的 KPI。每行都基于已抓取的架构或市场结构证据。

[CP006, CP026, CP037, CP039, CP040, CP041]
FP003: 护城河 / 准备度 KPI

评分卡展示最可能决定 Zama 差异化能否顶住替代方案的几个维度。

只有团队规模是直接观察到的数字;序数评分来自分析师对已保留证据的判断。

[CP007, CP008, CP009, CP010, CP037, CP038]

3.5 图表

Chapter 04

04财务情况

4.1 收入模式、变现面与代币经济

Zama 公开可见的经济模型比典型单产品 SaaS 公司更宽,但也远没那么明确。 协议上线前最清晰的商业化证明来自 TechCrunch 2024 年 3 月访谈。Rand Hindi 当时称,Zama 已在六个月前开始商业化,签下超过 $50 million 合同价值,对部分加密客户收取代币,对使用私有链的银行按交易收费。 这些表述意味着商业模式以企业许可和使用量定价为中心,而非自助订阅。 协议上线后发布的官方来源又加了一层变现:代币被设计用于加密、解密、跨链和运营商参与;公开拍卖和 CoinList 材料则描述了一次融资事件,清算价 $0.05,最终付费金额 $44 million。 这一点很重要,因为代币销售不应与运营收入混淆。条款明确否认股权、债务或收入分成权利,因此代币更像网络融资加未来效用经济,而不是经常性软件收入的直接替代品。[CI007, CI014, CI015, CI016, CI017, CI020]

收入来源表
来源机制单位当前值 / 状态质量尽调要求
企业技术许可区块链和 AI 客户将 Zama 密码技术用于商业场景合同价值 / 许可费官方称已向数十家公司授权;未公开收入拆分若合同续签,耐久性可能不错;但目前确认收入未披露索取客户数、ACV、合同期限和收入确认政策
银行 / 私有链定价面向私有区块链客户按用量收费按交易计费CEO 向 TechCrunch 提到;未公开价目表能证明定价可随用量扩展,但缺少合同支撑索取实际费率表、最低承诺和毛利率桥接
加密货币客户定价部分加密货币客户以代币支付以代币计价的商业费用CEO 向 TechCrunch 提到;未公开实际换算法说明能做加密原生变现,但折合法币后的收入质量不清楚索取以代币计价的合同样本和财库政策
协议效用费用加密、解密、跨链桥和运营者参与都需要代币协议代币效用官方条款定义了效用,但未公布实际费用量可能变成耐久的网络收入,但公开材料里还看不到现金流索取费用表、销毁 / 铸造政策和运营者经济模型
代币销售所得2026 年 1 月公开拍卖和平台销售拍卖清算价和中标出价中标出价最终支付 $44 million;属于融资,不是经营收入增强资本底盘,但不能证明经常性变现将财库所得与经常性软件或协议收入拆开
伙伴主导分发Dfns、Kinexys、T-REX 生态路径管线 / 分发渠道伙伴牵引可见,但未公开收入战略上重要,但公开层面尚未量化变现索取管线转化率、付费试点和分成条款

本表把经营性变现入口和融资事件分开。列入代币销售所得,是因为它影响资本充足性,不是因为它应计入经常性收入。

[CI007, CI015, CI016, CI017, CI029, CI030]
定价 / 变现表
项目价格 / 单位 / 合同标价与实际折扣 / 未知项来源 / 含义
B 轮股权融资$57 million,估值 >$1 billion已观察到的融资事件详细条款未披露已被公开交叉印证;说明投资人有兴趣,但不说明收入质量
A 轮股权融资$73 million,估值大约在 $300 million 后段至 $400 million已观察到的融资事件详细条款未披露TechCrunch 提供早期估值曲线锚点
公开拍卖清算价每枚代币 $0.05已观察到的公开销售结果中标分配和财库处理未完全公开有市场价格信号价值,但不是经常性价格实现
公开拍卖最终支付金额$44 million已观察到的公开销售结果会计处理和资金进入哪个实体未公开对资本有利,但应与软件收入分开
CoinList 底价和销售形式$55 million FDV 底价;$100 最低购买额;USDT / USDC 出资类似标价的公开销售机制二级市场表现和后续定价波动大更适合理解代币经济,不适合证明企业变现
加密货币客户商业定价部分客户用代币支付管理层表述,不是价目表交易量和代币会计处理不清楚暗示区块链原生变现,但可比性弱
银行 / 私有链商业定价按交易收费管理层表述,不是价目表未公开费率表或最低承诺指向用量经济,但毛利路径仍不透明
企业保密产品未留存官方价目表已审官方来源未发现标价折扣、支持和许可结构未知定价尽调缺口仍然很大

本表混合列出融资价格、代币销售机制和经营性定价证据,因为 Zama 公开经济信息横跨这三类。表中特意标明哪些条目属于融资,而不是经常性变现。

[CI001, CI005, CI015, CI016, CI023, CI025]
FI001: 收入模型桥接图

公开证据支持一条路径:先靠开源采用和企业试验,再走向授权和 token 效用经济;token 销售收入单列为融资。

仅为定性桥接。公开来源指出了收入入口,但没有披露当前结构或已确认收入贡献。

[CI007, CI014, CI015, CI016, CI029, CI030]

4.2 牵引信号、单位经济代理指标与成本结构

公开牵引证据存在,但大多是代理证据,不是会计证据。 Zama 自己的关于页面显示 96 人和 37 名博士;TechCrunch 此前报道,2024 年团队 75 人、3,000 名开发者使用这些库;到 2025 年 6 月,Tech.eu 和 EU-Startups 报道开发者超过 5,000 名。 Dfns、Kinexys 和 T-REX 分别以不同方式强化机构需求故事:Dfns 可能分销到 400 多家企业客户;Kinexys 给 Zama 一个大型银行沙盒证明点;T-REX 公告则把这套栈接到 ERC-3643 代币化基础设施上。 但这些来源都没有披露确认收入、付费客户数、平均合同价值或贡献利润率。 成本侧也只部分可见。TFHE-rs 文档和官方融资报道强调,GPU、FPGA 和 ASIC 加速是规模化路径,这意味着计算和研究强度会持续。 仅员工数增长也意味着固定成本基础在上升。因此,公开证据支持需求和技术投入真实存在,却不能搭出一条从产品使用到毛利润的清晰公开桥梁。[CI008, CI009, CI010, CI011, CI012, CI013]

单位经济模型表
指标数值 / 空值置信度重要性尽调要求
已签合同价值(2024 年引用)$50M+ 合同价值显示协议上线前已有需求,但不是确认收入索取已开票收入、回款和剩余履约义务
开发者牵引2024 年 3,000 名开发者;到 2025 年 5,000+开源采用的漏斗顶部信号不错索取活跃开发者数、生产部署数和付费转化率
员工数2024 年 75 人;2026 年关于页面为 96 人成本基础增长的核心代理指标索取按职能划分的薪酬和按地区划分的烧钱速度
机构分发触达400+ 个 Dfns 企业客户显示切入金融业的潜在渠道杠杆索取客户重叠、管线转化和在线付费部署
概念验证成熟度Kinexys 沙盒 PoC;T-REX 保密性集成说明与机构场景相关,但不能证明经常性收入索取付费试点状态和合同扩张路径
毛利率判断密码技术和基础设施能否带利润扩展,离不开它索取许可、协议和支持业务之间的毛利率桥接
CAC / 回本周期评估伙伴主导企业销售动作的商业化效率,离不开它索取按渠道拆分的获客成本和转化数据
收入集中度用于检验是否依赖少数试点或战略客户索取头部客户收入占比和合同期限

空值是刻意保留的。公开来源只给出牵引代理指标,没有给出完整承销模型所需的私营公司单位经济数据。

[CI008, CI011, CI012, CI013, CI017, CI033]
FI002: 单位经济性桥接图

公开记录说明需求可能如何转化为收入,但缺少闭环所需的成本和利润率数字。

该图有意保持定性,因为公开记录给的是机制证据,而不是计算 CAC 或利润率桥接所需的私营公司会计数字。

[CI013, CI017, CI019, CI033, CI034, CI035]
FI003: 财务估计区间(除注明外,单位:USD M)

公开资料可支撑融资和 token 销售经济性的区间,但无法支撑当前收入、烧钱或现金区间。

累计融资行是区间,因为独立来源同时使用欧元和美元口径的总额。当前收入或现金跑道没有公开区间,因为底层会计输入缺失。

[CI001, CI003, CI005, CI021, CI023, CI043]

4.3 资本充足性、法律实体与融资依赖

融资侧看,标题数字很强,流动性细节很弱。 2024 年 Series A 轮和 2025 年 Series B 轮形成了可见的股权基础;2026 年 1 月公开拍卖带来 $44 million 中标出价,并以 11,103 名独立竞价者显示强市场胃口。 官方和独立报道一致认为,Series B 轮用于资助主网上线、生态增长和持续 R&D。 这足以判断 Zama 抽象意义上并不缺资本,却不足以判断今天的现金跑道。 当前现金余额、月度烧钱速度和预期现金跑道仍未出现在保留的公开记录中。 法律实体图谱也重要。代币销售条款指向 Zama Switzerland AG 作为代币发行方,法国登记记录则指向 ZAMA SAS 作为运营实体。 公开来源没有把公司间资金流、许可安排或代币销售所得是否可供运营公司使用讲透明。 因此,财务能力方向上偏正面,但仍难以承销。[CI001, CI002, CI003, CI004, CI005, CI006]

资本充足性表
资本来源 / 信号金额 / 状态重要性资金用途 / 依赖披露缺口来源
A 轮(2024 年 3 月)$73M;估值接近 $400M奠定早期机构融资基础支撑商业化和招聘扩张当前剩余现金未披露TechCrunch
B 轮(2025 年 6 月)$57M;估值 >$1B协议上线前的关键股权补强主网、生态采用、研究和扩张没有资产负债表桥接或融资后现金官方 + Tech.eu + The Block + EU-Startups + Financial IT
公开拍卖所得(2026 年 1 月)最终支付 $44M;承诺 $118.5M增加流动性和市场验证支撑代币上线和网络经济,但所得去向未完全公开实体层面的财库可用性不清楚官方拍卖 + CoinList
当前账上现金现金跑道的核心承销输入没有它就无法判断偿债能力能撑多久未公开披露缺失
月度烧钱速度用来把融资额换算成现金跑道没有它就无法测试悲观情景下的融资依赖未公开披露缺失
现金跑道月数用来判断下一轮融资触发点当前融资强度仍可能掩盖短期过桥需求未公开披露缺失
法人实体拆分法国运营公司加瑞士代币发行方影响资金停在哪个实体,以及谁支撑运营可能让财库和收入解读变复杂公司间资金流不透明INPI / Pappers / 代币条款

本表聚焦当前承销相关性,不还原历史融资全时间线。这里的融资事实只来自财务章节内的本地论点,且只使用本地来源引用。

[CI001, CI003, CI005, CI006, CI023, CI028]
FI004: 资本强度 / 现金流地图

Zama 的现金叙事主要由股权和 token 销售流入主导,但流出端仍部分藏在算力、研究和实体结构不确定性后面。

资金流仅表示方向。公开来源支持这些流入和可能用途的存在,但不支持当前资金库余额或月度现金流出。

[CI004, CI019, CI023, CI028, CI043, CI044]

4.4 财务结论与尽调阻碍

因此,财务结论是混合的。 Zama 有可信的商业化证据、可信的机构兴趣,以及对一家深密码学公司而言异常强的融资支持。这些都是真正的正面因素。 但保留来源仍让核心承销问题悬而未决。登记和备案来源提供实体、资本和持续经营通知,却没有提供足以做下行情景分析的一组公开财务报表。 公开媒体来源提供签约合同价值、开发者采用和融资估值标记,但没有提供确认收入、年经常性收入(ARR)、毛利率、现金、烧钱速度或现金跑道。 反向记录不算灾难性,但有意义:Societe.com 保留了 2021 年净资产跌破资本一半后的继续经营通知;代币市场评论则强调完全解锁风险、估值缺口观感和波动的二级市场表现。 结果是,这家公司的增长与融资叙事,比收入质量和当前偿付能力证据链有更充分证据。 本章建议因此停留在“需求证据丰富、会计证据偏薄”。[CI039, CI040, CI041, CI042, CI043, CI046]

公开财务缺口表
缺失的私营公司指标对结论的影响当前公开状态具体尽调路径
确认收入 / ARR无法验证变现规模和质量留存公开来源没有披露当前确认收入或 ARR索取分收入流数据和过去 12 个月确认收入
毛利率卡住密码技术和基础设施经济性的承销判断留存公开来源没有披露毛利率索取收入成本桥接和基础设施成本分摊
现金余额卡住偿债能力和现金跑道分析未公开披露索取最新管理层资产负债表或财库备忘录
月度烧钱速度和现金跑道卡住下一轮融资时点分析未公开披露索取月度烧钱速度、悲观情景和最低现金目标
实际企业定价无法比较标价机制与实际变现只看到被引用的模式:部分客户用代币支付,银行按交易收费索取合同、折扣表和账单样本
客户数和集中度无法评估是否依赖少数大型试点或伙伴未公开披露索取客户细分和头部账户占比
公司间现金流无法看清代币销售所得是否资助运营留存来源同时出现瑞士发行方和法国运营公司索取财库结构和转移协议
经审计的公开财务报表无法做传统承销和契约式审查有注册摘录,但没有足以支撑承销的公开报表集索取经审计财务报表或投资人报告包

本表解释了为什么本章只能方向性判断资本可用性。公开信号真实存在,但会计资料仍然很薄。

[CI038, CI039, CI041, CI042, CI043, CI049]

4.5 图表

Chapter 05

05产品与技术

5.1 产品面已覆盖协议轨道、开发者库和面向用户的保密金融应用

Zama 已不再像一家披着公司外衣的单一密码学库。 到 2026 年,公开产品面覆盖面向公链的保密协议、保密智能合约框架 FHEVM、用于低层加密计算的 TFHE-rs、用于加密 AI 和分析工作流的 Concrete 与 Concrete ML,以及更新的 SDK 和用于屏蔽、桥接、质押、拍卖的协议应用。 这种宽度很重要,因为客户工作流很明确:开发者可以从库层开始,构建者可以把保密代币接入既有 Web3 应用,机构则可以使用预构建封装器和应用,而不必从零接线 FHE 原语。 最强证据是官方文档,加上 GitHub 和包页面:Zama 暴露的是实际开发者入口,而不只是发布白皮书。 需要注意的是,几个面向买方的用例仍依赖 Zama 运营的界面和封装器,因此日常实践中平台看起来更像一体化系统,而不是完全去中心化。[CE001, CE002, CE003, CE004, CE005, CE019]

产品模块 / 资产矩阵
模块 / 资产主要用户当前状态 / 成熟度差异化关键尽调缺口
Zama Protocol机构和 dApp 开发者主网已上线,包含封装合约、质押、跨链桥和保密金融定位给现有链增加保密性,而不是另起一条孤立隐私链生产延迟和故障处理的独立证据薄弱
FHEVMSolidity 开发者当前核心框架,代码库、文档和发版节奏都在线让加密状态与公开 EVM 状态和熟悉工具链共存性能多大程度依赖 Zama 运营的协处理器,公开基准测试还没给答案
TFHE-rsRust 工程师和协议团队稳定且持续发版的开源库纯 Rust TFHE,支持 CPU、GPU、HPU 后端,并提供 C 和 WASM 接入公开代码库文本仍把侧信道缓解描述为即将推出
Concrete / Concrete MLML 工程师和重隐私 AI 团队当前开源工具,已有演示和包分发让加密 ML 更接近 scikit-learn、PyTorch 和 Rust 流水线相比技术深度,商业客户引用偏少
Zama SDK + React SDK钱包、交易所和 dApp 团队仍是测试版,但公开文档齐备,现在可集成用 ERC-20 风格抽象和前端绑定藏起 FHE 复杂度测试版阶段意味着 API 稳定性和支持负载仍需直接验证
官方封装合约 + 注册表代币发行方和应用团队主网和测试网已可用开发者无需部署自定义封装,就能使用标准代币的保密版本封装合约治理和升级控制需要审查
投资组合 / 质押 / 跨链桥应用代币持有人和运营者当前面向用户的协议应用说明 Zama 能交付工作流,不只是密码原语应用可用性已有公开证明;企业支持义务还没有
开发者计划和生态漏斗初创公司、模板开发者和实验团队2026 年活跃把从实验到可投产应用的入门路径搭起来参与计划不等于耐久商业采用

本矩阵把当前可见的产品界面与路线图愿景分开;成熟度反映截至 2026-05-28 公开抓取材料所显示的状态。

[CE001, CE003, CE006, CE011, CE019, CE031]
工作流 / 用例表
用户任务当前工作流Zama 方案可衡量 / 可见收益关键限制
发行或封装保密代币部署代币,再配链下隐私权宜方案ERC-7984 封装合约和注册表余额和转账金额从转换后开始加密封装合约治理和监管访问规则仍需审查
构建保密智能合约自定义密码集成或单独隐私链FHEVM,配 Solidity 指南和 SDK 支持开发者仍能贴近熟悉的 EVM 工作流重计算仍要经由协处理器和配套服务
运行机构私密转账或结算许可制系统或可见的公开转账协议加委托解密和合规钩子加密执行,同时向托管方或监管方选择性披露企业 SLA 和事件响应的公开证据有限
执行加密 ML 推理或微调将数据移入可信环境,或推理前先解密Concrete ML 加 TFHE-rs 互操作加密 ML 流水线可留在 Python 和 Rust 生态内面向更重模型,延迟和成本仍然显著
发起私密拍卖、薪酬或代币分发链下批处理或公链透明性协议应用、拍卖流程和生态案例展示过去在公链上难以落地的新应用类别很多案例仍由伙伴主导或社区构建,还不是广泛客户铺开
把保密性接入现有钱包或 dApp UX重建自定义加密逻辑TypeScript 和 React SDK 包面向开发者的明文抽象降低 FHE 学习曲线SDK 仍是测试版,还需要真实用例里的开发者验证

收益来自公开产品主张和可见应用界面;不等同于经审计 ROI 或 SLA 支撑的性能。

[CE004, CE019, CE021, CE024, CE031, CE032]
FE002: 客户工作流 / 运营流程

构建者或机构通常如何从机密性需求走到由 Zama 支撑的应用上线。

不同产品从不同入口开始,但所有公开工作流都依赖加密、宿主链协调、链下计算和许可式揭示路径。

[CE004, CE019, CE024, CE031, CE032, CE033]

5.2 运营模式是分层的:公链保密合约、链下协处理器,以及 MPC 支撑的密钥控制

公开证据足够具体,能勾勒真实架构,而不是模糊的隐私故事。 FHEVM 是开发者直接接触的智能合约框架,但它只是其中一层。 这套栈还包括宿主链合约、网关、执行加密计算的链下协处理器、KMS 连接器、中继器,以及阈值 MPC 密钥管理设计。 这很关键,因为 Zama 明确想保留公链可组合性,同时把 FHE 中昂贵的部分移出基础链。 协议文档和第三方解读都描述了 handle、委托解密、封装器,以及用于密钥生成和访问控制的链上协调。 结果比纯研究原型更面向生产,但也意味着产品从链下运营商、专用计算和编排服务继承了真实依赖风险。 因此,投资者应把 Zama 看作一个有多重运动部件的保密运营层,而不是一个可以孤立评估的单一库。[CE006, CE007, CE008, CE009, CE010, CE011]

技术 / 运营架构表
层级 / 组件作用依赖风险
宿主链执行保密合约,记录状态转换Ethereum 及兼容 EVM 链底层链经济模型和宕机仍会影响系统
FHEVM 合约层定义加密类型、ACL、包装器和符号执行钩子Solidity 工具链和 Zama 库API 或版本频繁变动可能影响下游应用
网关 / 执行器协调请求、事件、验证和解密流程协议智能合约和运营方交易会成为延迟和事故处理的控制点
协处理器网络在链下跑高负载加密计算专用算力和诚实运营方执行运营复杂度和硬件成本不低
KMS / TKMS MPC 层持有分布式密钥份额,支持阈值解密独立 MPC 伙伴和安全协同阈值假设和伙伴在线率很关键
中继器 / SDK 层让应用加密输入、提交调用,并读取解密输出前端库和托管基础设施依赖托管中继器,可能压缩自托管灵活性
注册表 / 包装器映射受支持保密资产和规范合约地址协议治理和代币包装器维护升级或注册表失误可能波及多个应用

架构表同时纳入链上和链下组件,因为公链保密性要靠两侧一起跑通。

[CE006, CE008, CE009, CE010, CE024, CE028]
FE001: 产品架构图

Zama 的公开架构从面向用户的应用一路下探到加密计算和密钥管理核心。

这是基于协议文档、repo 结构和发布说明综合得出的分析性架构图,并非供应商正式发布的图示。

[CE006, CE008, CE009, CE024, CE028, CE033]
FE003: 关键依赖图

关键外部和内部依赖,决定 Zama 能否把机密计算交付成产品。

依赖图混合了协议模块和运营依赖,因为买方风险同时来自代码设计和服务执行。

[CE024, CE028, CE033, CE038, CE039, CE040]

5.3 发布节奏和开发者信号很强,但最大性能跃升仍主要由公司自行发布

异常密集的公开发布证据支撑了产品故事。 Zama 持续围绕 TFHE-rs、FHEVM、Concrete 和 Concrete ML 发布详细更新;开源仓库、文档、包注册表和开发者计划文章也显示,公司确实在争夺开发者心智。 TFHE-rs v1.4 和 Concrete ML v1.9 尤其重要,因为它们把 Zama 的隐私叙事绑定到 GPU 吞吐量、HPU 延迟和加密 ML 工作流的可量化变化上,而不是泛泛承诺。 协议测试网更新还指向 1.2 million 笔加密交易、19,000 个保密合约、120,000 个活跃钱包,以及 20 多个合作伙伴在构建。 这些指标是有意义的方向性信号,但仍压倒性来自公司自报,并与显著更高吞吐量、保密质押扩展等前瞻路线图主张并列存在。 证据支持真实执行速度和真实开发者采用,但尚未证明持久企业级经济性,或在持续生产负载下经独立审计的吞吐量。[CE013, CE014, CE015, CE016, CE017, CE018]

路线图 / 发布 / 开发阶段表
日期 / 阶段功能 / 里程碑状态含义来源
2025-07TFHE-rs、Concrete、Concrete ML 和 FHEVM 月度发布汇总已发布显示公司按多产品节奏推进,而非只押单一协议2025 年 7 月发布汇总
2025-12-31协议主网上线已发布将保密层从测试网推入生产环境开发者计划第 1 季
2026-01-21Zama Auction 主网上线已发布证明已有真实价值在场的 FHE 原生应用跑通开发者计划第 1 季
2026 测试网候选发布13 节点 MPC 网络,解密性能提升 10x已发布在更大规模扩张前强化去中心化和运营就绪度协议测试网更新
2026 当前SDK 测试版和 React 绑定已发布测试版降低钱包、交易所和 dApp 集成摩擦协议更新 / Grafa
2026 当前USDC、USDT、WETH、BRON、ZAMA、tGBP、XAUt 官方包装器已发布为保密代币使用提供标准入口协议更新
路线图保密质押、保密收益和智能体支付路线图可能把买方价值从包装器和转账扩到更广场景协议更新
路线图借助 GPU 和 ASIC 提升吞吐路线图若被验证,可能成为机构级负载的关键能力主页 / 公告

各行把已发布事项和前瞻性路线图主张分开;后者仍需独立验证。

[CE015, CE021, CE025, CE026, CE029, CE031]
FE004: 产品成熟度 / 能力图

截至 2026 年 5 月下旬,公开证据显示的 Zama 主要能力模块成熟度。

成熟度标签是基于公开产品证据、发布节奏和可见部署深度的分析判断。

[CE019, CE021, CE027, CE031, CE032, CE035]

5.4 信任控制在改善,但可靠性和许可仍制造尽调摩擦

相比许多加密基础设施项目,Zama 披露的信任和运营细节更多。 公开材料描述了 13 节点 MPC 设计、链上分布式密钥生成、面向受监管读者的委托解密、规模很大的首发审计投入,以及拆分核心服务的实时状态页。 这些都是有意义的正面因素,因为产品要赢得机构工作负载,必须同时具备保密性和可控可见性。 但可靠性图景并不完美。抓取时主网状态指标看起来很强,至少一个测试网服务仍显示明显更弱的可用性。 对一个仍在扩展栈的平台来说,这种分化可以接受;但它也意味着,单靠公开证据还不能断言韧性已达到银行级。 商业上,公司开源姿态是真实的,但代码仓库也明确显示,商业生产使用需要专利许可。 这可能帮助变现,却也可能相较更容易自托管或独立商业化的基础设施,拖慢生态采用。[CE030, CE033, CE038, CE039, CE040, CE041]

信任 / 质量 / 合规表
控制项 / 指标状态范围缺口
13 节点 MPC 网络公开材料称已在最新测试网 / 候选发布架构上线阈值密钥生成和解密需要第三方证明运营方独立性和历史阈值可用性
分布式密钥生成公开材料称已上链并成功完成密钥生成和恢复流程未看到公开事故日志或恢复演练证据
审计计划公开材料称协议组件累计约 70 个审计周密码学设计和实现审查未获取到打包公开的审计报告合集
委托解密协议更新中的当前功能托管方、合规方或监管方读取权限具体策略模板和滥用控制需要客户验证
主网服务可用性抓取时状态页显示 MPC 在线率 99.964%、协处理器在线率 100%运营可靠性公开资料未提供正式 SLA 或长期在线率历史
商业许可代码库称商业使用需要专利许可企业部署和开源复用相比更宽松的技术栈,可能拖慢中立生态采用

本表区分控制项的公开披露和独立验证的运营证据;公开描述有方向性价值,但不能替代尽调材料。

[CE028, CE030, CE033, CE038, CE039, CE050]

5.5 结论:Zama 具备真实产品和技术优势,但最强公开证明仍来自自身运营足迹

产品与技术层面的核心投资逻辑可信。 Zama 有连贯的栈、真实文档、已上线协议应用、有意义的开源页面,以及加密智能合约工具已走出实验室的公开证据。 最强差异化在架构上:Zama 不要求用户迁移到新的隐私链,而是试图把保密能力加到已经存在的网络和工作流上。 如果团队能保持高抽象、低运营复杂度,这会成为强采用切口。 主要风险也清楚。公开证据仍高度依赖公司发布的性能数字、路线图目标和 Zama 运营服务;测试网到主网的可靠性差距可见;商业许可叠加也意味着对每个下游构建者而言,开放并非无摩擦。 归纳来看,产品和技术有投资价值,但下一步尽调应是围绕可用性、延迟和生产支持做真实客户访谈,而不是继续看更多协议营销。[CE002, CE024, CE025, CE027, CE038, CE039]

5.6 图表

Chapter 06

06客户情况

6.1 Zama 可见客户群集中在金融基础设施,而不是广泛企业席位部署

公开客户图景是真实的,但很集中。 Zama 最强证明来自保密能力作为前提的重基础设施细分:机构代币化轨道、OTC 交易台、企业钱包基础设施、浏览器,以及开发者构建的金融应用。 具名例子包括:J.P. Morgan 旗下 Kinexys 用于沙盒金融隐私工作流;T-REX Ledger 用于基于 ERC-3643 的代币化;GSR 用于保密 OTC 结算;Dfns 用于企业钱包分销;Bron 用于保密薪资;Blockscout 则围绕保密代币提供浏览器支持。 这是一组有意义的买方、用户和渠道组合,但不像典型 SaaS 客户基础那样有大量披露席位数的合同。 Zama 实际销售或赋能的是更广泛受监管金融工作流中的基础设施节点。 承销含义是:单次集成的客户价值可能很高,但由于公开证明基础仍锚定少数旗舰关系,集中度和伙伴依赖比原始 logo 数更重要。[CU001, CU004, CU005, CU006, CU007, CU009]

客户分层表
客群买方 / 用户 / 付费方使用场景规模 / 证据收入 / 战略价值缺口
机构代币化基础设施资产发行方、过户代理、RWA 运营方保密 ERC-3643 转账和报告$32B+ ERC-3643 基础资产;Apex 通过 T-REX 的目标为 $100BT-REX 集成放量后战略价值高未披露合同经济性或收入分成
机构交易台做市商、OTC 交易台、资金运营方保密 OTC 结算和转账规模控制GSR 提供具名证明证明机构级交易场景可落地目前只有一个标杆证明;重复交易量未披露
银行和金融基础设施创新方数字资产战略团队和合规团队私募基金认购、拍卖、DvP、加密 KYC / AMLKinexys 沙盒提供具名证明顶级可信度信号仍披露为概念验证(PoC),不是大范围铺开
钱包和托管基础设施钱包平台、托管方、金融科技运营方保密钱包、策略和受监管交易流Dfns 覆盖 400+ 企业客户;Bron 已在主网上线单个集成带来强分发杠杆尚不清楚有多少 Dfns 客户启用 Zama 功能
浏览器和工具层区块浏览器、合规工具、开发者基础设施在不暴露余额的情况下展示保密代币Blockscout 和生态伙伴提供具名证明提升可发现性和工作流兼容性工具分发不等于付费使用
开发者应用层初创公司、开发者、黑客松团队薪资、钱包、法律协议、AI 支付轨道季度计划,以及 880 个黑客松项目中的 51 个积累未来客户和渠道管线很多应用仍处早期,并明确未经审计

客户分层聚焦已有公开证据的买方群体;同时区分渠道 / 基础设施杠杆与直接经常性席位客户披露。

[CU004, CU005, CU006, CU007, CU009, CU011]
客户增长 / 采用轨迹表
指标数值日期 / 期间来源可信度含义缺失分母
已处理加密交易1.2M+2025-2026 测试网期间显示网络已有实质使用和实验有多少延续为主网收入
活跃钱包 / 地址120K+2025-2026 主网上线前活动说明用户和开发者参与度已超出小范围早期测试钱包不等于付费客户
已部署保密合约19K+2025-2026 主网上线前活动说明开发者确实在部署加密逻辑看不到留存下来的活跃合约
构建应用和集成的伙伴20+2025-2026 更新显示生态扩张高度依赖伙伴未披露伙伴深度或变现
可触达的 Dfns 企业客户400+2026 当前一个集成就能显著拓宽渠道触达激活率和付费转化率未披露
ERC-3643 代币化资产基础$32B+2026 当前给 T-REX / Zama 打入大型合规优先资产池的路径实际有多少份额使用 Zama 保密能力
Apex 对 T-REX Ledger 的承诺到 2027 年 6 月 $100B2026 公告若兑现,可能形成未来渠道规模承诺面向未来,不是当前使用量
基于 Zama 搭建的黑客松项目51 / 880Q1 2026显示实验性漏斗顶部很强黑客松后的存活率未知

这些是公开采用和渠道指标,不是经审计的收入、客户数或续约披露。

[CU006, CU007, CU012, CU020, CU026, CU030]
FU001: 客户旅程图

在最强的公开 Zama 客户和合作伙伴案例中,从机密性需求走向更广采用的典型路径。

[CU005, CU009, CU011, CU013, CU017, CU042]

6.2 具名客户证明在机构试点和基础设施集成中最强

对一家隐私基础设施创业公司来说,具名证明基础异常具体,但覆盖多个成熟度层级。 Kinexys 仍被描述为沙盒里的概念验证,因此它作为一线机构验证很有价值,却不等同于生产合同。 T-REX 更像平台级集成,背后接着一条与 ERC-3643 和 Apex Group 代币化雄心绑定的大型下游资产管线。 GSR 是最干净的生产式证明,因为 Zama 和 GSR 都描述了一笔在 Ethereum 上、KYCd 对手方之间完成的保密 OTC 交易。 Dfns 重要,是因为它能把 Zama 功能传播到 400 多家企业客户;Bron 重要,是因为在 Ethereum 主网上执行保密薪资,比泛泛合作新闻稿更具体。 Blockscout 也有用,因为它显示加密资产出现在面向用户的基础设施中,而不是埋在协议演示里。 主要限制在于,许多证明仍停留在试点、基础设施或渠道层,因此它们传递技术可信度的速度,快于传递经常性商业耐久性的速度。[CU005, CU006, CU007, CU009, CU010, CU011]

具名客户证明表
客户 / 伙伴客群部署 / 使用场景生产 / 试点结果 / 证明质量局限
J.P. Morgan 旗下 Kinexys银行 / 数字资产创新沙盒中的基金认购、盲池交易、DvP 结算、加密 KYC / AML试点 / 概念验证(PoC)强机构品牌背书,且披露了具体工作流细节仍未被描述为大范围生产部署
T-REX LedgerRWA 基础设施面向 ERC-3643 代币化资产的原生保密层集成 / 平台部署大规模下游资产基础和 Apex 承诺释放规模信号Zama 只是更大技术栈中的一层;直接变现路径不清楚
GSR机构做市商Ethereum 上首笔保密 OTC 交易接近生产环境的已完成交易抓取材料中最具体的机构执行证明一笔标志性交易不能证明持续交易量或留存
Dfns企业钱包基础设施面向银行、金融科技、托管、RWA 和支付客户的保密钱包与交易栈当前集成可能触达 400+ 企业客户和多个使用场景的渠道客户激活和收入捕获未公开
Bron钱包 / 资金运营在 Ethereum 主网上用 cUSDT 做保密薪资生产使用场景展示的是已上线的公司财务工作流,不只是演示证据仍局限在一个钱包和一个标杆工作流
Blockscout浏览器 / 工具原生支持 ERC-7984 保密代币当前工具集成说明保密资产开始进入终端用户基础设施工具存在感弱于支出或合同证据

各行覆盖截至 2026-05-28 在抓取材料中出现的主要具名证明,并刻意保留试点、集成、生产标签,不把它们压成同一个信心层级。

[CU005, CU006, CU007, CU009, CU011, CU012]
按具名部署拆分的公开证据新鲜度与证明质量
名称最新抓取到的证据证据质量生产成熟度结果具体性尽调含义
J.P. Morgan 旗下 Kinexys2026 年 Zama 文章中高试点验证力度强,但仍不是已披露的生产合同
T-REX Ledger2026 年 Zama 与 Chainwire 报道平台集成视为战略渠道证据,并核实直接变现
GSR2026 年 GSR 与 Zama 公告生产交易当前最强的机构执行证据;索取重复交易量
Dfns2026 年 Zama 与 Dfns 页面当前集成有价值的分发渠道,但缺少功能启用数据
Bron2026 年薪资主网上线文章中高生产工作流用例证据不错,但客户基础窄
Blockscout2026 年 Zama 与 Blockscout 文章中高当前工具支持有用的 UX 兼容信号,但商业信号较弱
Raycash / 社区应用2026 年生态与计划页面测试网 / 早期可算作管线,不是核心经常性客户证据
Builder 计划获奖者2026 年 Zama 计划页面实验性不错的前瞻指标;明确未经审计

本表把证据新鲜度、结果具体性与商业耐久性拆开;后者大多仍未披露。

[CU002, CU003, CU005, CU009, CU011, CU017]
FU003: 客户证明矩阵

按新鲜度、结果具体性、生产成熟度和投资判断价值为具名公开证明排序。

[CU005, CU006, CU009, CU011, CU016, CU017]

6.3 采用数据偏生态:网络活动和构建者吞吐可见,合同经济性不可见

与同阶段多数基础设施协议相比,Zama 拥有更多公开采用信号;但对于直接承销客户质量而言,这些信号类型并不理想。 协议和第三方报道指向超过 1.2 million 笔加密交易、超过 120,000 个地址或钱包,以及一套广泛的构建者、封装器和应用生态,它们正在准备使用或已经使用主网。 Shielded 报告补充说,880 个黑客松参赛项目中有 51 个由 Zama 构建的项目;开发者计划也显示,Zama 持续激励保密金融和工具团队。 这些支持真实漏斗顶部和真实开发者兴趣。 但它没有披露其中多少团队转化为付费生产客户,多少使用量在主网而非测试网,或集成是否随时间扩张。 实务中,采用曲线在生态宽度和基础设施嵌入上强于披露的、收入支撑的账户扩张。 这仍然是正面因素:对一个想成为公链保密层的协议来说,深生态分销可能比原始席位数更适合作为早期信号。[CU019, CU020, CU021, CU022, CU024, CU025]

FU002: 采用 / 部署漏斗

Zama 的漏斗顶部活动如何从构建者和标准工作,推进到具名部署和近生产证明。

[CU020, CU022, CU024, CU025, CU026, CU042]

6.4 因公开留存指标缺失,耐久性证据大多是间接的

客户章节最大弱点不在于 Zama 有没有具名证明,而在于这些关系是否粘性强、能扩张、且收入足够实质。 抓取到的公开来源没有披露净留存率(NRR)、总留存率(GRR)、流失、续约率、合同长度或头部客户收入占比。 最好的公开耐久性代理指标是渠道强度和运营证据:Dfns 宣传 99.95% 可用性和可识别参考客户;GSR 将保密性描述为重大机构痛点;协议栈已有上线应用,且超过一半流通 ZAMA 被质押。 这些信号有用,但不能替代商业留存指标。 同一问题也出现在旗舰 logo 上:Kinexys 仍是概念验证,T-REX 是带有下游承诺的结构性集成,许多社区应用明确标注未经审计。 正确读法是:客户质量有希望,但尚未被承销。 未来尽调应聚焦续约路径、首次集成后的账户扩张,以及伙伴主导分销能否转化为经常性使用,而不是一次性上线事件。[CU014, CU019, CU023, CU041]

留存 / 重复使用 / 满意度表
指标 / 代理指标数值 / 状态客群可信度尽调要点
净收入留存率(NRR)未公开披露全组合索取按机构集成队列、开发者转生产队列拆分的 NRR
总收入留存率(GRR)未公开披露全组合索取头部集成的 GRR、客户留存和续约排期
流失 / 合同期限未公开披露标杆客户索取合同期限、续约权和生产扩张时点
Dfns 可靠性代理指标Dfns 声称 99.95%+ 在线率渠道合作伙伴 / 企业钱包基础设施验证其中有多少在线率专门适用于保密代币工作流
社区粘性代理指标流通 ZAMA 中 50%+ 已质押协议用户 / 运营方区分代币持有人信心和企业客户耐久度
使用持续性代理指标主网上线前 / 早期主网叙事中的 1.2M+ 加密交易和 120K+ 钱包网络参与者拆分月活钱包、重复构建者和付费生产应用
证据质量代理指标很多社区项目明确标注未经审计开发者应用索取关键应用引用的安全审查或生产就绪证明

公开耐久性证据主要是代理指标。这些行都不能替代真实续约、流失或扩张指标。

[CU014, CU019, CU023, CU035, CU041]

6.5 扩张上行真实存在,但证明集中在少数旗舰渠道,集中度风险同样真实

Zama 的扩张逻辑很直观。 一旦客户或渠道集成保密代币和委托解密,同一保密层就能延伸到结算、托管、薪资、代币化资产、资金库运营和受监管 DeFi。 Dfns 尤其重要,因为一次集成就打开通往数百家企业客户的路径。 T-REX 通过 ERC-3643 资产流提供类似乘数;GSR 和 Kinexys 则展示 Zama 如何嵌入具体机构交易工作流。 风险在于,这些强项本身也制造集中度。公开证据集中在一小组旗舰关系、少数标准机构,以及重伙伴的市场进入动作上。 竞争不确定性也仍可见:独立报道仍把 FHE 描述为与 ZK 和许可制替代方案并列的竞争者,即使支持性文章也承认延迟取舍或法律可执行性争议。 归纳来看,Zama 有可信的先落地再扩张动作;但在直接收入、续约和扩张数据可得之前,投资者不应把生态兴奋度误读为分散化客户集中度。[CU015, CU031, CU032, CU036, CU037, CU038]

扩张和集中度风险表
扩张驱动 / 集中度风险类型影响尽调路径
Dfns 渠道杠杆扩张驱动一个集成就能把 Zama 功能触达多个金融垂直领域的 400+ 企业客户按 Dfns 客户队列索取激活、收入分成和扩张指标
T-REX / ERC-3643 资产管线扩张驱动因素可能让 Zama 接入大量合规优先的代币化资产要求提供实际运行的保密资产规模,而非整个标准口径的资产总额
机构工作流深度扩张驱动因素GSR 和 Kinexys 证明 Zama 能嵌入结算和交易流程,不只是消费端应用索取首次上线或概念验证后的重复使用证据
标杆客户集中集中度风险公开证据集中在少数几个名称上,标杆客户集中度可能高于预期索取前 10 大收入或使用量集中度,以及对每个旗舰集成的依赖
伙伴驱动的 GTM 模式集中度风险许多证据来自渠道、标准或基础设施伙伴,而非直接、经常性终端客户澄清每个旗舰关系里谁向谁付费,毛利率落在哪一层
性能与法律取舍争议集中度风险独立报道仍把 FHE 放在 ZK 和许可制替代方案之间,视为有争议的路线之一验证真实客户是否愿意接受延迟和法律模式取舍
早期应用质量集中度风险未经审计的社区应用如果按生产部署计算,会夸大客户广度核心客户评分表应只纳入经审计或有独立引用的生产实施

扩张路径可信,但加速采用的伙伴集中度,也抬高了依赖风险。

[CU012, CU020, CU030, CU031, CU036, CU042]

6.6 图表

Chapter 07

07风险

7.1 监管与法律边界

Zama 的风险首先不是纯密码学,而是法律架构。公开网站由法国 SAS 运营;代币销售、拍卖和拍卖专用隐私声明,则由位于 Zug 的 Zama Switzerland AG 发布。这个拆分本身不构成问题,但投资者必须同时追踪两套法律足迹、多项管辖法律条款,以及几套彼此独立的条款。网站条款把争议置于法国法和巴黎上诉法院之下;代币销售链条又叠加了瑞士公司和制裁筛查义务。代币拍卖和销售条款明确排除受制裁人士、被拒绝方以及 AML 不合作司法辖区居民;销售隐私声明则称 Zama 会收集钱包地址,以及来自 Sumsub 的 KYC 数据。这个方向与公司“隐私可以和合规共存”的说法一致,但也意味着协议可投资性高度取决于跨司法辖区 AML、制裁、准入和披露控制能否执行到位。 外部监管正在收紧,正好压在 Zama 想商业化的领域。ESMA 将 MiCA 描述为一套覆盖加密资产活动透明度、披露、授权和监管的统一制度;EU travel rule 则把发起人和受益人可追踪性延伸到加密资产转移。FATF 和 FinCEN 从另一个角度给出同样判断:私密或加密的价值转移,并不会取消可追踪性预期。Zama 的选择性披露设计是一种缓释,因为它试图保留监管者或审计方访问入口;但公开材料尚未展示经过监管检验的合规落地,也未披露牌照状态。因此,任何押注其快速拿下受监管金融客户的投资论点,剩余风险都很高。[CR001, CR002, CR003, CR004, CR005, CR006]

监管 / 法律风险登记表
规则 / 案件 / 文件司法辖区状态可能性严重性缓释措施剩余暴露尽调路径
MiCA 授权、披露和监管义务欧盟已生效,实施措施仍在推进选择性披露设计和受监管金融定位在公开合规姿态披露前仍为高索取关于 MiCA 边界、白皮书义务和任何 CASP 依赖的法律备忘录
欧盟 Travel Rule 对加密资产转移的可追溯要求欧盟已按 Regulation (EU) 2023/1113 生效可编程访问控制,加上代币销售流程中的 KYC 准入加密转移仍需合规共享信息,因此仍为高审查协议对发起人 / 受益人数据交换及交易对手工具的支持
拍卖 / 代币销售中的制裁与 AML 筛查瑞士 / 欧盟 / 美国已写入拍卖和代币销售条款中高地理围栏、名单筛查、Sumsub KYC、用户陈述执行质量尚未经公开检验,因此为中索取筛查供应商、误报和升级处理流程指标
网站和代币销售隐私义务法国 / 瑞士 / 欧盟已发布公开通知;代币销售通知于 2025 年 12 月更新中等隐私通知、GDPR 框架、瑞士 / 欧盟数据处理通知钱包、ID 和链上数据在多个界面被收集,因此为中索取 DPO 备忘录,覆盖留存、删除和合法性基础映射
SAS 网站运营方与瑞士发行方之间的跨辖区法律拆分法国 / 瑞士当前结构中等网站、代币销售、拍卖和质押分别有条款投资人必须尽调多个法律实体,因此为中逐实体梳理协议、网站、销售和质押争议的责任归属

各行按剩余暴露排序,仅概括 2026-05-28 检索到的公开文件;不对非公开牌照或监管沟通作判断。

[CR001, CR002, CR004, CR005, CR006, CR007]
FR001: 风险热力图

剩余风险集中在监管、运营方集中度和商业化不透明度。

[CR006, CR011, CR015, CR018, CR024, CR027]

7.2 运营与安全执行

协议已经不只是概念:Zama 有公开测试网使用数据、完成的去中心化密钥生成仪式、上线的质押,以及至少一笔主网工资交易。即便如此,从技术上亮眼的发布走到机构级可靠性,路径仍然很窄。测试网更新披露了超过 120 万笔加密交易、超过 19,000 份保密合约、120,000+ 活跃钱包和 20+ 伙伴;FHEVM v0.9 则被定位为首个主网候选版本,并把解密性能提升 10x。这些都是强进展信号,但也暴露出这套技术栈仍然高度依赖性能。DKG 仪式中每方传输了 10.5 TB 数据,协议也仍依靠阈值 MPC 和相对较小的运营商集合,来维持密钥管理去中心化。 主网上线没有消除实施风险。Zama 的上线后更新称全栈已经上线且可集成,但 SDK 仍处于测试版,委托解密也只是刚发布的新原语。质押上线时有 18 家运营商,分为 5 个 FHE 节点和 13 个 KMS 节点;超过一半流通供应量已经质押。这改善了经济对齐,但也意味着,只要有限基础设施伙伴中出现运营中断、密钥管理故障或协调失败,就可能同时冲击用户信任和估值。第一笔保密工资交易是真正的里程碑,但单笔工资交易还不能替代长期公开 正常运行时间记录、广泛应用多样性或披露的事故指标。[CR017, CR018, CR019, CR020, CR021, CR022]

运营 / 质量 / 安全风险登记表
失效模式可能性严重性缓释成熟度剩余暴露未解决缺口
FHE 性能或解密延迟在生产规模下失效正在改善但尚未成熟;已披露 10x 解密增益中高没有独立披露的吞吐量 / 延迟 SLA
13 节点架构中的阈值 KMS 或 DKG 协调失败低中极高仪式证据强,且未披露 DKG 失败未公开披露灾备或密钥轮换操作手册
SDK 测试版不成熟,拖慢伙伴集成或引发实施错误SDK 已存在并上线,但仍标为测试版中高没有公开企业端到端使用测试版工具迁移的案例
生产用例从拍卖、薪资扩展后,主网可靠性下降据报道拍卖没有停机;质押已上线没有公开正常运行时间看板或事故复盘档案
5% 发行和持续变化的 40/60 奖励分配下,运营者激励错配中等代币经济学和流动性质押设计已上线没有关于运营者盈利能力或流失风险的公开证据

严重性基于公开披露的架构和上线里程碑,而非任何私人审计结果或内部正常运行时间数据集。

[CR017, CR018, CR019, CR020, CR021, CR022]
FR002: 风险传导图

监管和架构失败会传导为合作伙伴转化放慢、需求走弱和估值压缩。

[CR021, CR024, CR025, CR027, CR030, CR032]

7.3 伙伴与商业依赖

Zama 的商业叙事高度依赖外部验证的机构用例。这是优势,因为公司不是抽象地卖隐私,而是把保密性绑定到工资、钱包、OTC 交易、代币化 RWA,以及 JPM 组织的隐私研究上。但这也是集中风险,因为大量证据仍由试点、PoC 或伙伴推动。J.P. Morgan 的 Project EPIC 明确把隐私、身份和可组合性列为代币化金融的前提;但被引用的成果是一篇论文加概念验证,不是规模化生产项目。GSR 公告同样被定位为概念验证。T-REX 和 Dfns 更有商业含金量,因为它们更接近机构分发:T-REX 称 ERC-3643 下已有 $32B 上链,Apex 承诺到 2027 年 6 月目标 $100B;Dfns 则称由 Zama 驱动的保密基础设施可提供给 400+ 企业客户。 这些伙伴叙事仍留下执行缺口。投资者还没有公开证据能判断 PoC 到经常性合同的转化率、伙伴收入贡献、客户集中度或标准化 SLA。运营商名单很强——Fireblocks、Ledger、Figment、DFNS、OpenZeppelin、Etherscan 等——但同一名单也把声誉和运营依赖集中在相对小的基础设施专家圈子里。只要其中少数标杆伙伴降低 Zama 优先级,协议的商业可信度就可能比密码学本身更快下滑。[CR024, CR030, CR031, CR032, CR033, CR034]

伙伴 / 依赖风险登记表
依赖交易对手角色集中度失败情景严重性缓释措施剩余暴露
机构隐私验证J.P. Morgan Kinexys / Project EPIC 项目代币化金融隐私和身份的参考 PoC叙事集中度高PoC 未推进为可重复的生产部署把机构参考案例扩展到银行研究试点之外
钱包和企业分发Dfns面向 400+ 企业客户的保密钱包基础设施中高Dfns 优先投入其他轨道,或采用仍然浅层扩展到更多钱包和托管集成中高
RWA 代币化分发T-REX Ledger / Apex 生态ERC-3643 流程的保密层中高已宣布的大额资产承诺没有转化为在线保密流程扩大发行方范围,并发布真实交易案例中高
机构市场结构证明GSR保密 OTC 执行概念验证PoC 停留在象征层面,没有经常性交易量中等拿下更多 LP、交易所或资金管理用例
协议运营者集合Fireblocks、Ledger、Figment、OpenZeppelin、DFNS 及同类机构KMS/FHE 节点和生态可信度一个或多个运营者退出、作恶或无法扩展极高维持超额抵押质押,扩大运营者集合,并记录替换路径

本登记表把概念验证认可与生产分发拆开,避免高估伙伴关系强度。

[CR024, CR030, CR031, CR032, CR033, CR040]
人员 / 执行风险登记表
角色 / 职能依赖或缺口可能性严重性缓释措施尽调路径
创始人 / CEO 领导力Rand Hindi 仍是融资、协议愿景和机构叙事的公众代表技术可信度深,生态关系可见审查继任计划和商业领导梯队
从研究到商业化的转换96 人、37 名博士的团队意味着密码学深度强,但规模化 GTM 运营的公开证据有限与 Dfns、T-REX、GSR 和 OpenZeppelin 的伙伴关系带来分发杠杆索取销售领导组织图、配额和转化指标
跨境法律协调法国 SAS 网站运营与瑞士 AG 代币发行带来多实体执行要求中等各个界面已有独立法律文件索取实体责任矩阵和内部升级路径
合规运营公开文件显示有 KYC、制裁和隐私义务,但没有说明由谁运营、服务水平如何第三方 KYC 供应商加正式法律条款索取合规人员配置、审核员覆盖和审计节奏

人员登记表聚焦运营深度的公开证据缺口,不猜测未披露的员工流失。

[CR001, CR004, CR005, CR037, CR038, CR042]
FR003: 依赖关系图

Zama 要把隐私基础设施转成可持续采用,得靠运营方、渠道和机构验证者。

[CR024, CR030, CR031, CR032, CR033, CR034]

7.4 财务包袱与否决条件

Zama 的融资数据很亮眼——2025 年 Series B 融资 $57M,估值超过 $1B——但最可投资的问题不是公司有没有融到钱,而是监管、商业化和代币分发风险,能否在公司必须证明持久变现之前被压下来。公开拍卖数据令人鼓舞,因为它显示兴趣强,且未报告停机;但也显示需求质量受退款影响很大:$118.5M 承诺额最终只有 $44M 由中标方支付。这是有用的价格发现,不是稳定长期需求的证明。公司也称协议上线前已向数十家公司授权技术,但公开收入、留存、正常运行时间 和转化指标仍然缺位。 实际投资含义是,Zama 的风险不能只靠密码学打分。会打破投资论点的事件可以监控:制裁、AML 或 MiCA 合规失败;重大安全或解密事故;运营商集中度变得不稳定;或伙伴试点无法转成生产使用。反过来,如果 Zama 披露公开审计、经常性机构收入、面向监管的合规工作流,以及重复的生产交易量,剩余风险会显著下降。在此之前,协议应因真实上线进展获得认可,但证据基础仍更偏向技术可行性,而不是规模化商业耐久性,因此剩余风险仍高。[CR026, CR027, CR028, CR035, CR036, CR039]

缓释与终止标准表
风险可监控触发因素阈值 / 事件行动含义
监管适配破裂公开的 MiCA / Travel Rule 合规问题、制裁违规,或因合规担忧导致伙伴下线任何与保密控制相关的正式执法行动或重大准入暂停视为受监管金融采用逻辑破裂
运营架构破裂主网事故、解密中断或运营者集合不稳定新增节点前反复停机、密钥管理事件失败,或可见的运营者流失暂停投资判断,直到可靠性证据改善
试点到生产转化停滞拍卖、薪资或零散伙伴试点之外,没有更多生产案例12 个月内没有可重复的企业交易增长或具名生产客户把判断从基础设施领导者下调为研究资产
需求质量恶化上线后二级代币需求或质押参与走弱质押参与持续低于当前流通量多数水平,或上线后出现可见流动性压力降低对代币经济学支撑安全模型的信心
商业证据持续不透明没有公开收入、正常运行时间、审计或客户留存披露披露改善前又进入下一轮融资周期要求保护性条款,或推迟投资

终止标准有意设为可监控,并绑定可观察的法律、运营、需求和披露事件。

[CR023, CR026, CR027, CR029, CR036, CR040]

7.5 图表

Chapter 08

08估值

8.1 估值背景与价格纪律

公开证据不给出一个干净的 Zama 单一价格,而是给出一个区间。最近一个明确披露的股权数据点,是 2025 年 6 月估值超过 $1B 的 Series B。Premier Alternatives 的二级市场式数据指向约 $840M。Alpha Drops 显示 2026 年 1 月公开代币销售估值为 $550M;Zama 自己的拍卖页面又补充了细节:承诺额 $118.5M、实际支付仅 $44M、出清价 $0.05,且超额认购很重。这些数据点并不完全可互换——代币出清、二级市场指示和私募股权估值,是流动性和控制权不同的工具——但合在一起说明一件重要事:市场尚未给 Zama 定出一个显而易见的价格。 这种模糊性很重要,因为 Zama 仍没有公开披露收入、ARR、利润率或留存。在这种背景下,不能像给一家财务 KPI 清晰的私有 SaaS 公司那样,用同样信心给 Zama 估值。公开价值论据建立在三点上:机密计算类别领先、协议确实上线的证据,以及来自运营商、拍卖和机构渠道的早期需求信号。这足以支持继续尽调和不低的战略溢价;但如果没有未披露商业化数据作为条件,还不足以承销任何高于 $1B 的价格。[CV001, CV002, CV003, CV004, CV005, CV006]

建议摘要表
维度取值决策含义
建议继续研究只有商业和法律尽调补上最大证据缺口后才推进
信心主网进展真实,但公开财务披露仍太薄,难以形成高信心
风险评级监管复杂度和收入证据缺失拉大结果分布
估值立场超过 >$1.0B 时偏高;~$550M-$840M 的公开证据区间更合理入场价格应反映财务披露不完整
持有 / 重估触发因素只有披露费用消耗、收入和企业转化数据后才重估不要仅因技术新颖就支付溢价

取值仅基于截至 2026-05-28 抓取到的公开证据;不假设代币市场与股权市场数据点在经济含义上相同。

[CV001, CV002, CV003, CV027, CV029, CV039]
FV002: 估值敏感性

公开定价锚点横跨代币市场出清价和私募轮定价,入场估值因此成为核心决策变量。

[CV001, CV002, CV003, CV029, CV037, CV040]

8.2 投资论点与反论点

Zama 的正面论点很直接。协议已经不只是白皮书:质押已上线,密钥生成已完成,上线后技术栈可以集成,第一笔保密工资已在 Ethereum 主网执行,保密更新称 Dfns 可以把加密交易开放给 400+ 企业客户。市场顺风也真实存在。Ripple/BCG 和其他分析来源继续把代币化资产描述为万亿美元级建设;J.P. Morgan 和 Figment 也都认为,隐私、身份和可组合性是机构代币化金融的前提。如果保密性确实是公链金融缺失的那一层,Zama 就是最可信、且公开可见地在补这一层的协议之一。 反论点同样清楚。公开财务披露仍太薄,无法判断 Zama 是否找到了可重复的经济引擎,还是主要靠技术里程碑支撑估值。与隐私相关的资产仍面临监管折价,公司自己的公开定价点也从 $550M 到超过 $1B 不等。团队精英化,但规模仍不到 100 人;公司必须同时扩展协议运营、法律合规、机构销售和开发者工具,单个员工承受的执行负担很高。换句话说,这是一家产品论据可能走在财务证明前面的公司。[CV007, CV008, CV009, CV010, CV011, CV012]

投资逻辑 / 反向逻辑表
维度投资逻辑反向逻辑
市场需求代币化金融越来越需要隐私、身份和可组合性代币化资产 TAM 材料不能证明 Zama 能捕获经济价值
产品验证主网质押、DKG、薪资支付和 SDK 上线,显示出真实交付速度SDK 仍是测试版,公开正常运行时间 / 事故证据仍稀少
分发Dfns 及其他运营方 / 合作伙伴渠道能加快企业触达公开机构证据大多仍以 PoC、合作伙伴或公告为主
定价支撑多个公开定价点拼出可投资区间同一组定价点也说明市场尚未形成一个可信价值
护城河质量可编程保密性与传统隐私币定位拉开差异合规落地一旦卡住,监管折价仍会压缩任何隐私相关协议
资本基础Series B 轮为生态建设提供现金跑道已融资本不能替代收入、毛利率或留存证据

这组反向逻辑有意围绕价格和证据敏感性展开,而不是泛泛的质量批评。

[CV001, CV010, CV011, CV012, CV013, CV014]
FV001: 建议逻辑

当前建议靠真实产品验证和品类顺风支撑,但财务披露薄、监管折价仍在抵消。

[CV010, CV011, CV015, CV016, CV027, CV031]
FV004: 投资 KPI

Zama 的品类位置和早期验证较强,但财务可见性和定价清晰度偏弱。

[CV010, CV015, CV016, CV018, CV027, CV031]

8.3 可比对象与情景

可比集合无法给出整齐公式,但仍有信息量。Fireblocks 是乐观的私募市场锚点:成熟的机构加密基础设施平台,据报道二级市场式估值 $8B,已融资 $1B。Digital Asset 是更冷静的企业账本锚点,在 Premier Alternatives 上约 $380M。Arqit 提供公开市场警示样本:一家上市密码学公司,可以在极高收入倍数上交易,却仍只有几十万美元的当期收入。Zama 介于这些两端之间。它的类别动能强于小众上市加密安全微盘股,但披露的分发广度远少于 Fireblocks。 因此,情景分析比点估值更自然。牛市情景里,Dfns 式分发加上机构隐私需求,转化为可重复交易量和费用销毁,从而验证或超过最近一轮私募股权估值。基准情景里,Zama 继续发货,维持强质押参与度,并扩大生产用例,但投资者仍需要在财务证据不完整的情况下控制仓位。熊市情景里,监管继续高压,与隐私相关的代币仍被折价,协议难以把技术可信度转化为可持续经济性,合理价值回落到代币销售或二级市场区间。今天的公开证据最支持基准情景,而不是牛市情景。[CV019, CV020, CV021, CV022, CV023, CV024]

乐观 / 基准 / 悲观情景表
情景概率关键假设指示性估值结果回报逻辑 / 含义
乐观20%企业渠道转化,费用消耗变得可见,质押保持强劲,新的生产级金融应用上线>$1.2B 至 ~$1.6B可支撑按上一轮私募估值附近或略高价格入场
基准50%主网采用增长,但公开收入披露仍不完整,监管仍复杂~$700M 至 ~$1.0B只支持在严格价格纪律下加入观察名单 / 选择性入场
悲观30%监管折价持续,试点未能扩规模,代币需求在上线后走弱~$400M 至 ~$600M公允价值向代币销售或折价老股交易式水平收敛
概率加权按当前证据加权,而不是管理层叙事~$770M 中点指向继续研究,而不是买入

估值结果是基于公开定价区间和商业化证据搭出的启发式情景锚点,并非贴现现金流输出。

[CV002, CV003, CV010, CV013, CV028, CV029]
可比估值表
可比对象指标估值 / 状态相关性局限
Zama 公开代币销售2026 年 1 月代币销售估值隐含估值 ~$550M;融资 $44M最接近协议需求实时市场出清的数据点代币销售在经济属性上不等同于普通股
Zama 老股交易式标记Premier Alternatives 私营公司快照市场隐含估值 ~$840.3M最接近公司股权的公开老股交易式标记方法论和流动性深度并不完全透明
FireblocksPremier Alternatives 私营公司快照市场隐含估值 ~$8.0B;融资 $1.0B展示规模化机构加密基础设施可以达到的估值Fireblocks 的托管和交易覆盖面宽得多
Digital AssetPremier Alternatives 私营公司快照市场隐含估值 ~$380M;融资 $457.2M可作为企业账本基础设施的对照架构、成立时间和市场周期经历不同
Arqit QuantumYahoo Finance 公开市场快照市值 ~$280.35M;EV/Revenue ~233.58x带有当前市场定价的公开密码学基础设施基准收入基数太小,倍数不稳定,也不能直接迁移

表中混合了股权、老股交易式和代币市场参照,因为拥有公开财务数据的纯 FHE 可比公司很少;每一行都明确写出自身局限。

[CV002, CV003, CV019, CV020, CV022, CV023]
FV003: 估值 / 回报区间

现有公开证据支持较宽估值区间,基准情形低于上一次披露的股权融资轮。

[CV002, CV003, CV029, CV037, CV038, CV040]

8.4 建议与最终尽调

正确建议是继续研究,置信度中等,风险评级高。Zama 已明显跨过值得严肃关注的门槛:它交付了主网组件,组织了有声誉的运营商,展示了实际运行 用例,并以独角兽估值融资。但投资者不应把强技术可信度误认为已被证明的估值支撑。如果价格高于最近披露的 >$1B 轮次,公开记录太薄,不足以下买入判断。在约 $550M 到 $840M 之间,证据更有吸引力,因为入场价更好地反映了收入未披露和商业化问题仍未关闭。 最终尽调清单因此要窄而实用。决策者需要经审计或管理层认证的收入和费用销毁 数据、毛利率或服务成本证据、生产客户集中度,以及对股权结构表和代币包袱机制的清晰视图。法律侧需要确认,MiCA、travel-rule、制裁和隐私义务,已经转成运营工作流,而不是只停留在使用条款措辞。如果这些披露结果正面,估值就能从区间框架扩展为更常规的增长型基础设施承销模型。如果不能,价格纪律应压过热情。[CV026, CV027, CV028, CV029, CV030, CV031]

投资逻辑破裂和止损触发表
触发因素阈值对投资逻辑的传导行动含义
商业证据迟迟不出现到下一次重大融资或金库事件时,仍没有收入 / 费用消耗披露估值仍由叙事驱动,而不是经济性驱动不要在估值区间中段以上做投资测算
监管折价扩大针对隐私相关资产和选择性披露设计的新限制或交易所反弹代币和股权倍数被压缩将基准情景移向悲观区间
合作伙伴转化停滞除当前上线和合作伙伴公告外,没有新增生产案例企业分发逻辑明显弱化下调确信度,要求更低入场价
网络安全性走弱质押参与度或运营方质量从当前占流通量多数和 18 家运营方阵容的水平恶化安全性和可信度溢价快速压缩重新评估代币和股权投资测算
股权结构表或悬置风险爆雷代币解锁、排放或金库行为的支撑力度低于暗示水平公开定价区间高估可持续价值暂停资金部署,直到股权结构表清晰度提升

每个触发因素都应靠公开或管理层提供的证据监测,而不是靠主观叙事更新。

[CV009, CV010, CV031, CV032, CV036, CV037]
最终尽调问题表
主题缺失证据为什么重要负责人 / 尽调路径
收入质量费用消耗、协议收入、企业合同价值和客户集中度缺少这些,估值无法跳出基于区间的启发式判断管理层财务包,加上从代币活动到收入的 KPI 桥接
单位经济性毛利率、运营方补贴负担、每笔保密交易的服务成本用来检验费用消耗能否跑赢排放和基础设施成本财务主管 / FP&A 尽调会
法律准备度MiCA / 旅行规则 / 制裁操作流程和法律意见用来判断机构采用是否被合规摩擦卡住外部律师备忘录和合规工作流演示
股权结构表和代币悬置金库持仓、解锁时间表、权利以及股权 / 代币互动用来理解稀释,以及代币定价是否会脱离股权价值法务 / CFO 股权结构表审查
客户证据除薪资支付外的具名生产部署,以及重复交易量用来区分概念验证和可持续商业化客户背调电话和产品使用仪表盘

这些问题刻意收窄到能决定投资的事项;相比更多市场规模报告,这些缺口更能快速改变建议。

[CV026, CV027, CV036, CV042]

8.5 图表

免责声明

本报告为截至 2026-05-28、仅基于公开信息生成的 AI 尽调摘要。报告不构成投资建议;作出任何投资决策前,还应补充管理层尽调、法律审阅和一手财务材料。

证据索引

结论
编号陈述可信度来源
CO001 Zama SAS is a French simplified joint-stock company registered in Paris under SIREN 879243319 and headquartered at 8 rue du Sentier, 75002 Paris. SO003, SO004
CO002 The company registered its activity in late 2019 even though management and partner materials describe Zama as founded in 2020. SO004, SO005, SO006
CO003 Rand Hindi is the current CEO of Zama and is named as the website publisher representative in the legal notice. SO003, SO021
CO004 Jeremy Bradley is identified publicly as publication manager and COO, and INPI records list Jeremy Bradley as Directeur Général. SO003, SO004
CO005 Zama was founded by Rand Hindi and Pascal Paillier. SO005, SO007, SO021
CO006 Zama describes itself as an open-source cryptography company building fully homomorphic encryption solutions for blockchain. SO001, SO002
CO007 The current website positions Zama around the Zama Protocol, the $ZAMA token, staking, bridging, developer tooling, and enterprise confidentiality use cases. SO001, SO002
CO008 Zama’s about page reports a workforce of 96 people, including 37 PhDs across 26 nationalities. SO001
CO009 AWS describes Benoît Chevallier-Mames as Zama’s VP of Product Engineering. SO005
CO010 TechCrunch reported that Zama had a team of 75 in March 2024 and planned to use Series A proceeds to hire more engineers. SO020
CO011 Zama raised a $73 million Series A in March 2024 led by Multicoin Capital and Protocol Labs. SO006, SO020
CO012 Series A participants also included Metaplanet, Blockchange, VSquared, Stake Capital, Portal Ventures, and strategic founders such as Juan Benet, Gavin Wood, Anatoly Yakovenko, Julien Bouteloup, and Tarun Chitra. SO006
CO013 TechCrunch reported that Zama had previously raised $8 million across pre-seed and seed rounds, bringing total funding to $81 million at the time of the Series A. SO020
CO014 Zama announced a $57 million Series B in June 2025 at a valuation above $1 billion. SO007, SO021, SO022
CO015 Pantera Capital and Blockchange Ventures co-led the Series B round. SO007, SO021, SO023
CO016 Independent coverage described Zama as the first unicorn in the fully homomorphic encryption sector after the Series B. SO021, SO022
CO017 Post-Series-B total funding exceeded $150 million. SO021, SO022, SO024
CO018 Management said the Series B was designed to add strategic blockchain investors ahead of mainnet and token launch rather than simply maximize capital raised. SO007, SO021
CO019 The Series B announcement said Zama already licenses its FHE technology to dozens of companies across blockchain and AI. SO007
CO020 Zama positions the Confidential Blockchain Protocol as a cross-chain confidentiality layer on top of existing L1 and L2 chains rather than a new blockchain. SO007, SO014, SO015
CO021 The company’s public stack spans TFHE-rs, Concrete ML, FHEVM, SDKs, and the protocol documentation, showing a broad open-source product surface beyond a single application. SO006, SO014, SO016, SO017, SO018, SO019
CO022 GitHub shows the Zama organization with 3.7k followers and Paris listed as location. SO016
CO023 At fetch time, TFHE-rs showed about 1.6k GitHub stars and 324 forks. SO017
CO024 At fetch time, Concrete ML showed about 1.4k GitHub stars and 197 forks. SO018
CO025 At fetch time, FHEVM showed about 25.3k GitHub stars and 2k forks, far larger than Zama’s other public repositories. SO019
CO026 Zama’s about page says its investors have backed many of the most successful blockchain companies globally. SO001
CO027 Dfns says it joined Zama’s first MPC operator cohort alongside Figment, InfStones, LayerZero, Omakase, and Stake Capital. SO027
CO028 Dfns says Zama’s public testnet is live and describes Ethereum mainnet and TGE as the next rollout steps. SO027
CO029 The litepaper states Ethereum mainnet is live, other EVM chains are planned for H1 2026, and Solana support for H2 2026. SO015
CO030 Chainwire and Zama state that T-REX chose Zama as the default confidentiality layer for its ERC-3643 ledger. SO008, SO025
CO031 The T-REX materials cite $32 billion of tokenized assets on ERC-3643 and a target of $100 billion on Apex-backed infrastructure by June 2027. SO025
CO032 Zama announced a GSR collaboration focused on confidential institutional trade execution on Ethereum. SO009
CO033 Zama announced an OpenZeppelin partnership centered on audited confidential contracts for DeFi and digital assets. SO010
CO034 Zama announced that Kinexys by JPMorgan released a proof of concept built with Zama FHEVM. SO011
CO035 The official shielded report and Dfns operator post position Dfns as a pathway to encrypted transactions for hundreds of enterprise clients. SO013, SO027
CO036 AWS says Zama uses Hpc7a instances for heavy FHE workloads and Nitro Enclaves around key management, indicating the product still depends on specialized infrastructure to operate at scale. SO005
CO037 TechCrunch characterized homomorphic encryption as still lacking mass-market scalability and said the overall market remained small in early 2024. SO020
CO038 Cointelegraph framed institutional privacy infrastructure as an active competition among FHE, zero-knowledge systems, and permissioned-network designs. SO026
CO039 MEXC argued that Zama entered public-token price discovery with an aggressively priced private valuation and remained unproven at scale. SO030
CO040 CoinGabbar reported that ZAMA fell nearly 50% from its February 2026 debut while 11 billion total supply and 2.2 billion circulating supply created selling pressure. SO028
CM001 Public blockchains expose transactions, balances, and state transitions to everyone, creating a core confidentiality dilemma for sensitive financial workflows. SM001, SM002, SM003
CM002 Institutional users are reluctant to place proprietary trading logic, investor data, or sensitive balances on fully public ledgers. SM003, SM004, SM005
CM003 Zama’s target market is confidentiality on top of existing public chains rather than a separate privacy chain. SM001, SM002, SM010
CM004 A meaningful excluded set includes private chains, TEE-led models, zero-knowledge-only privacy stacks, and MPC-heavy designs that solve adjacent problems differently. SM001, SM005
CM005 The litepaper treats stablecoins as a leading blockchain payment use case with trillions in yearly volume. SM001
CM006 Forbes, citing a16z research, reported $4.5 trillion of stablecoin payment volume in Q1 2026. SM011, SM012
CM007 The same Forbes report said 2025 stablecoin transaction volume crossed $33 trillion and stablecoin supply reached roughly $320 billion. SM011
CM008 Forbes also noted that nearly two-thirds of Q1 2026 stablecoin payment volume originated from Asia. SM011
CM009 Crowdfund Insider, summarizing CoinGecko research, said tokenized real-world assets grew 256.7% from $5.42 billion at the start of 2025 to $19.3 billion by the end of Q1 2026. SM016
CM010 Tokenized U.S. Treasuries remained the largest RWA category at 67.2% share and crossed the $10 billion milestone in early 2026. SM016
CM011 Crowdfund Insider reported that tokenized commodities grew 289% to $5.55 billion and reached 28.7% of the RWA market in Q1 2026. SM016
CM012 Tokenized stocks and ETFs were much smaller than Treasuries and commodities but still reached roughly $487 million and $300 million respectively by the end of Q1 2026. SM016
CM013 RWA.xyz shows today’s tokenized-asset leaderboard is concentrated in credit, Treasuries, gold, and money-market instruments rather than a broad long tail of assets. SM017
CM014 Chainwire said ERC-3643 already secures about $32 billion of tokenized assets and Apex-backed infrastructure targets $100 billion by June 2027. SM004
CM015 BCG, Aptos Labs, and Invesco said tokenized fund AUM exceeded $2 billion in late 2024 and could reach more than $600 billion by 2030. SM015
CM016 Ledger Insights said McKinsey’s 2024 base case put tokenized assets below $2 trillion by 2030 with a $1 trillion to $4 trillion range. SM014, SM018
CM017 AssetTokenization.com summarized 2030 tokenization forecasts from roughly $1.9 trillion on the low end to more than $30 trillion on the high end. SM018
CM018 The litepaper treats as much as $100 trillion of financial assets as theoretically movable onchain, but that figure is a broad aspirational TAM rather than a near-term serviceable market. SM001
CM019 The most immediate buyer groups visible in retained sources are asset issuers or fund managers, wallet and custody infrastructure providers, DeFi application developers, and institutional trading or compliance operators. SM004, SM006, SM007, SM008
CM020 In issuer-led workflows, the payer and budget owner are usually the financial institution or asset manager that needs compliant tokenization and settlement. SM004, SM015, SM026
CM021 Dfns shows wallet and custody providers can become both buyers and operators because confidential transaction rails require MPC and enclave infrastructure as well as customer-facing distribution. SM006, SM009
CM022 Developer adoption cost falls when confidentiality can be added through existing Solidity tooling rather than a new language or new chain. SM001, SM007, SM010
CM023 The OpenZeppelin partnership is evidence that developer tooling and standards support are part of the go-to-market wedge, not just cryptography performance. SM007
CM024 The Kinexys proof of concept is evidence that bank and market-infrastructure experimentation exists, but it remains a proof point rather than proof of scaled production spend. SM008
CM025 Core adoption drivers include confidentiality for sensitive data, selective compliance disclosure, interoperability on public chains, and reduction of front-running or MEV. SM001, SM003, SM004, SM006
CM026 A major differentiator for Zama’s market story is that users do not need to bridge into a new chain to access confidentiality. SM001, SM002, SM003
CM027 The docs and fundraising materials say FHE performance improved more than 100x over five years but still relies on GPUs and future accelerators to reach higher throughput. SM001, SM010, SM024
CM028 TheBlock and later summaries place Zama near 20 transactions per second before the next hardware step to 100+ and eventually thousands per second. SM019, SM024
CM029 Cointelegraph shows institutions still face competing privacy models, including zero-knowledge systems and permissioned networks, so Zama is not selling into an uncontested category. SM005
CM030 Dfns and Figment-style operator narratives imply buyer adoption also depends on trusting a multi-party KMS and coprocessor network, not just a Solidity library. SM003, SM006
CM031 The market story depends heavily on regulatory comfort with selective disclosure, auditability, and KYC/AML controls rather than raw privacy alone. SM001, SM004, SM026
CM032 Forbes cited IMF research showing markets increasingly price stablecoins as a competitive threat to incumbent payment companies. SM011
CM033 Crowdfund Insider said clearer regulatory frameworks from 2024 to 2025 helped accelerate institutional participation in RWAs during 2026. SM016
CM034 CoinGabbar said ZAMA had fallen nearly 50% from debut and highlighted 11 billion total supply versus 2.2 billion circulating supply as a selling-pressure overhang. SM022
CM035 Phemex said the token sale raised about $121 million through a sealed-bid Dutch auction and the token launched with heavy early volatility. SM020
CM036 Blockchain.News argued that broken momentum and negative funding could send ZAMA toward $0.025, showing how speculative market structure can undermine enterprise messaging. SM023
CM037 MEXC argued that Zama’s public valuation should be discounted for unproven scale and the fully unlocked public-sale dynamics. SM021
CM038 The serviceable market today is narrower than any trillion-dollar TAM because visible production evidence clusters around payments, tokenized funds or RWAs, and a few institutional workflow pilots. SM004, SM015, SM016
CM039 Apex-backed T-REX, Dfns operator infrastructure, and OpenZeppelin tooling suggest the most realistic adoption path is ecosystem-by-ecosystem integration rather than instant horizontal ubiquity. SM004, SM006, SM007
CM040 Zama’s market can become large, but only if institutional pilots, developer tooling, and token economics reinforce each other faster than performance, governance, and volatility concerns accumulate. SM003, SM015, SM021, SM022
CP001 Zama positions its protocol as a confidentiality layer that sits on top of existing L1 and L2 chains rather than as a new blockchain. SP001, SP022
CP002 Zama says developers can build confidential applications with familiar languages such as Solidity and Python without needing cryptography expertise. SP001
CP003 Zama says deploying applications on the protocol is free and does not require an additional license from Zama. SP001
CP004 Zama says the protocol keeps transaction inputs and smart-contract state encrypted end to end, including from node operators. SP001
CP005 Zama says its FHE technology is already about 100 times faster than it was five years earlier. SP001
CP006 Zama says protocol throughput rose from roughly 0.5 transactions per second to more than 20 transactions per second and that GPU, FPGA, and ASIC acceleration are the path to higher throughput. SP001
CP007 Zama’s about page listed 96 people on the team at the 2026 run date. SP002
CP008 Zama’s about page listed 37 PhDs on the team. SP002
CP009 The Zama–Dfns announcement says Dfns integrated Zama’s confidential token standard for more than 400 enterprise clients across banking, fintech, custody, RWA, and payments. SP004
CP010 Zama reported a successful proof of concept in J.P. Morgan’s Kinexys Digital Assets sandbox, indicating institutional experimentation with its confidentiality stack. SP003
CP011 Fhenix markets itself as bringing fully homomorphic encryption privacy to Ethereum through its CoFHE engine. SP005
CP012 Fhenix explicitly says its privacy stack uses no trusted execution environments and no zero-knowledge workarounds. SP005
CP013 Fhenix says developers can add privacy with standard Solidity and Uniswap v4 hooks without moving to a new VM or rollup. SP005
CP014 Fhenix documentation says sensitive data remains encrypted throughout computation. SP006
CP015 Inco says it is a full-stack privacy layer for blockchains instead of a new blockchain. SP007
CP016 Inco says it works with MetaMask, Coinbase Wallet, Phantom, and mainstream Solidity, Rust, Hardhat, Foundry, and Anchor tooling across EVM and SVM networks. SP007
CP017 Secret Network documentation says it has offered private smart contracts on mainnet since September 2020. SP008
CP018 Secret Network says it uses Intel SGX, encryption schemes, and key management rather than FHE to deliver confidentiality. SP008
CP019 Secret Network’s site now pitches privacy-first AI and confidential-computing use cases, not only DeFi or privacy-preserving transfers. SP009
CP020 Oasis Sapphire is marketed as a confidential EVM environment with Ethereum Virtual Machine compatibility. SP010
CP021 Oasis Sapphire claims 99% or more lower fees than Ethereum for its confidential smart-contract environment. SP010
CP022 Oasis Sapphire claims 6-second finality and cross-chain bridges, making it a faster-integrating option for some EVM-native teams. SP010
CP023 Oasis broader documentation says trusted offchain logic runs in TEEs on a permissionless network of nodes. SP011
CP024 Aztec documentation says Aztec is a privacy-first Ethereum zkRollup that is not EVM compatible. SP013
CP025 Aztec documentation says private functions execute and prove on user devices and private state is stored as encrypted UTXOs. SP013
CP026 Aztec’s main site says centralized sequencers, backdoors, and TEEs do not meet its privacy bar. SP014
CP027 Aleo documentation describes Aleo as a platform for building private applications with its own programs, wallet, staking, and node-running workflow. SP015
CP028 Aleo’s bug bounty is live on HackerOne and Bugcrowd. SP016
CP029 Aleo said its bug bounty launched with an initial $500,000 reward pool and OFAC-compliant payout rules. SP016
CP030 Partisia documentation says its confidential-computing product uses MPC and also supports verifiable credentials. SP017
CP031 AWS Nitro Enclaves are hardened isolated VMs with no persistent storage, no interactive access, and no external networking. SP018
CP032 Azure confidential computing markets protected data-in-use processing as general infrastructure for deploying any application on confidential compute. SP019
CP033 Google Confidential Space says multiple parties can process sensitive data with an agreed workload while even the operator cannot access the data being processed. SP020
CP034 Arcium positions blockchain-based MPC confidential computation as easy to integrate with only a few lines of code. SP021
CP035 Figment says Zama adds a modular privacy layer to existing blockchains instead of replacing the base layer. SP022, SP001
CP036 Figment says Zama’s execution model depends on offchain FHE coprocessors plus threshold KMS and access-control lists. SP022
CP037 ChainSafe says 2026 institutional adoption is being driven by privacy architecture choices rather than by one universally accepted privacy stack. SP023
CP038 KuCoin’s 2026 market overview names Zama, Fhenix, and Inco among the projects making private blockchains practical. SP024
CP039 BlockEden’s 2026 comparison says no single privacy technology is universally best and that FHE, ZK, and TEE each carry distinct tradeoffs. SP025
CP040 Cloud confidential-computing products have enterprise procurement familiarity and general-purpose infrastructure advantages that blockchain-native privacy layers do not. SP018, SP019, SP020, SP023
CP041 Zama’s moat is strongest where a buyer wants pure cryptographic privacy on existing chains rather than trusted hardware or a new execution environment. SP001, SP022, SP025
CP042 Zama’s main competitive risk is that buyers may accept TEE-based or zk-based substitutes as good-enough privacy if they bring lower migration friction or stronger installed distribution. SP011, SP014, SP018, SP019, SP020, SP023, SP025
CI001 Zama raised a $57 million Series B at a valuation above $1 billion. SI001, SI017, SI018, SI023, SI024
CI002 The Series B was co-led by Pantera Capital and Blockchange Ventures. SI001, SI017, SI018, SI023, SI024
CI003 Public coverage around the Series B said total funding had risen to more than $150 million, with one euro-denominated article describing more than €129 million. SI017, SI018, SI023, SI024
CI004 Series B proceeds were described as funding mainnet launch, ecosystem adoption, research, and scaling confidential financial applications. SI001, SI023, SI024
CI005 TechCrunch reported that Zama raised $73 million in a March 2024 Series A at a valuation near the high end of $300 million to $400 million. SI016
CI006 TechCrunch reported that before the Series A, Zama had raised $8 million across pre-seed and seed rounds. SI016
CI007 Zama said it had been licensing its technology to dozens of companies across blockchain and AI before unveiling the protocol. SI001, SI018
CI008 Zama’s about page showed 96 people at the 2026 run date. SI002
CI009 Zama’s about page showed 37 PhDs on the team. SI002
CI010 Zama’s about page showed 26 nationalities on the team. SI002
CI011 TechCrunch reported Zama had a team of 75 in March 2024. SI016
CI012 TechCrunch reported around 3,000 developers were using Zama’s libraries in 2024. SI016
CI013 Tech.eu said more than 5,000 developers were using Zama’s libraries by June 2025. SI017, SI023
CI014 Zama documentation publicly shows a product surface spanning protocol apps, SDKs, TFHE-rs, Concrete, and Concrete ML rather than a single end-user product. SI008, SI009, SI010, SI011
CI015 TechCrunch quoted Rand Hindi saying some crypto customers pay Zama in tokens. SI016
CI016 TechCrunch quoted Rand Hindi saying banks using private blockchains pay Zama by transaction. SI016
CI017 TechCrunch quoted Rand Hindi saying Zama had signed north of $50 million in contract value within roughly six months of commercialization. SI016
CI018 TFHE-rs documentation says the library includes GPU and HPU accelerated backends. SI011
CI019 Official and media sources said Zama expected GPU, FPGA, and ASIC acceleration to move the protocol from roughly 20 transactions per second toward tens of thousands, implying continued infrastructure investment needs. SI001, SI018, SI022, SI024
CI020 The public auction ran from January 21 to January 24, 2026. SI003, SI019
CI021 Official auction results said Zama reached $121.3 million of total value shielded and $118.5 million of committed value across platforms. SI003
CI022 Official auction results said the sale executed 24,697 bids from 11,103 unique bidders. SI003
CI023 Official auction results said the clearing price was $0.05 and the final amount paid by winning bids was $44 million. SI003, SI019, SI025
CI024 Official auction results said 880 million tokens were sold and demand exceeded supply by 218 percent. SI003
CI025 CoinList said the total token supply was 11 billion, the allocated public-sale supply was 880 million tokens, and the FDV floor was $55 million. SI019, SI025
CI026 CoinList said the minimum purchase amount was $100 and purchases were funded in USDT or USDC. SI019
CI027 CoinList said the unlock schedule for purchased tokens was February 2, 2026. SI019
CI028 The token sale terms name Zama Switzerland AG, based in Zug, as the issuer of the Zama token. SI004
CI029 The token sale terms say the token is required for encryption, decryption, and bridging of data via the protocol. SI004
CI030 The token sale terms say the token is necessary to become an operator and may be delegated to operators. SI004
CI031 The token sale terms say the token does not represent debt, equity, or rights to future revenue or capital flows. SI004
CI032 The token sale terms require KYC and include sanctions and anti-money-laundering restrictions for buyers. SI004
CI033 The Dfns partnership made Zama’s confidential token standard available to 400-plus enterprise clients. SI005
CI034 Zama reported a successful proof of concept in the Kinexys Digital Assets sandbox. SI006
CI035 Zama’s T-REX announcement said more than $32 billion of assets already lived onchain under ERC-3643 and that Apex Group had committed to placing $100 billion on the T-REX Ledger by June 2027. SI007
CI036 INPI lists ZAMA SAS with SIREN 879243319 and an RNE registration date of 27/11/2019. SI013, SI012
CI037 INPI lists Zama’s activity code as APE 6201Z and states that the company purpose includes homomorphic-encryption technology. SI013
CI038 INPI shows capital social of €2,395.50 while Pappers and Societe.com show capital social of €2,372.33, so the public filing extracts are not perfectly reconciled. SI013, SI014, SI015
CI039 Pappers says no public financial information is available for ZAMA SAS. SI014
CI040 Societe.com records a continuation decision despite net assets having fallen below half of capital in 2021. SI015
CI041 The reviewed public record does not disclose current recognized revenue, ARR, cash balance, burn rate, or runway months. SI014, SI015, SI016, SI017, SI018, SI023, SI024
CI042 The reviewed public record shows enterprise licensing, transaction pricing, token utility, and partner-distribution economics, but no official enterprise list price card. SI001, SI008, SI016, SI019
CI043 Capital adequacy is directionally improved by the Series B and the $44 million public-sale proceeds, but the present cash bridge remains unverified because cash and burn are undisclosed. SI001, SI003, SI017, SI018, SI023, SI024
CI044 Official and media coverage frame the use of funds around mainnet launch, ecosystem growth, research, and scaling financial applications rather than around a disclosed path to near-term profitability. SI001, SI023, SI024
CI045 The token sale is financing rather than operating revenue because the token terms deny equity or revenue-share rights and sale materials focus on supply allocation, bidder demand, and clearing price. SI003, SI004, SI019
CI046 Cryptomaniaks argues that the full TGE unlock and VC-to-retail valuation gap make early public participation highly speculative. SI021
CI047 MEXC framed Zama’s launch as a shift from the privacy-coin narrative toward confidential-computation infrastructure. SI022
CI048 Figment says Zama’s design uses offchain FHE coprocessors, threshold key management, and access-control lists, which supports the product thesis but also adds operating complexity. SI020
CI049 Public filing and registry extracts provide entity detail but not a set of audited public financial statements suitable for underwriting. SI012, SI013, SI014, SI015
CI050 The shift from 75 employees in March 2024 to 96 people on Zama’s 2026 about page implies a rising cost base even before considering specialist cryptography compensation. SI002, SI016
CI051 Partnerships with Dfns, JPMorgan/Kinexys, and T-REX are credible institutional GTM signals, but none of the retained public sources disclose conversion of those relationships into recurring revenue. SI005, SI006, SI007
CI052 Because the operating company is ZAMA SAS in France while the token issuer is Zama Switzerland AG, public sources do not make intercompany cash flows or treasury availability transparent. SI004, SI013, SI014
CI053 ICO Analytics listed the token around $0.036 at fetch time, about 0.72x the $0.05 auction price, underscoring volatile aftermarket reception. SI025
CE001 Zama describes itself as an open-source cryptography company building fully homomorphic encryption solutions for blockchain. SE001, SE002
CE002 The Zama Confidential Blockchain Protocol is positioned as a confidentiality layer for any L1 or L2 rather than a new standalone blockchain. SE001, SE004
CE003 The public Zama product suite spans the protocol, FHEVM, TFHE-rs, Concrete, and Concrete ML. SE001, SE005
CE004 Zama’s public documentation includes quick-start guides, Solidity guides, SDK docs, protocol apps, and code examples. SE005, SE006
CE005 The about page lists 96 people, 37 PhDs, and 26 nationalities. SE002
CE006 FHEVM is presented as the core framework of the Zama protocol for confidential smart contracts on EVM-compatible blockchains. SE011, SE006
CE007 FHEVM publishes guarantees around end-to-end encryption of transactions and state, encrypted and public state coexisting onchain, and composability with existing dApps. SE011
CE008 The FHEVM repository exposes gateway-contracts, host-contracts, a coprocessor, a KMS connector, and end-to-end test infrastructure. SE011
CE009 FHEVM supports encrypted integer types up to 256 bits and is designed to offload FHE computation asynchronously to a coprocessor. SE011, SE019
CE010 The v0.6 fhEVM release added euint128, euint256, larger encrypted byte arrays, input proofs, and a coprocessor path that can run on Sepolia without changing the base chain. SE019
CE011 TFHE-rs is a pure Rust implementation of TFHE with Rust, C, and client-side WASM APIs. SE007, SE009
CE012 TFHE-rs documentation says the library includes GPU and HPU accelerated backends in addition to CPU support. SE007, SE009
CE013 The TFHE-rs GitHub page says the first stable version v1.0.0 shipped in February 2025 for the x86 CPU high-level API. SE009
CE014 TFHE-rs advertises compression, size-efficient public-key encryption, programmable bootstrapping, and exact integer arithmetic as part of the library feature set. SE009
CE015 TFHE-rs v1.4 claims all GPU operations are at least 2x faster than v1.3 and 64-bit division reaches up to 4x speedup on four or more GPUs. SE017
CE016 TFHE-rs v1.4 says keyswitching key size falls by half and FheUint64 operations improve by roughly 10 to 19 percent on CPU. SE017
CE017 TFHE-rs v1.4 claims integer-operation latency on HPU fell by up to 45 percent after algorithmic and clock-speed changes. SE017
CE018 TFHE-rs v1.4 says a new ERC20_SIMD instruction reached 87 ERC20 transfers per second on a single HPU. SE017
CE019 Concrete ML is positioned as an open-source privacy-preserving machine-learning framework built on top of Concrete. SE008, SE010, SE015
CE020 Concrete ML presents a scikit-learn-like interface and can also convert PyTorch models for FHE execution. SE010, SE015
CE021 Concrete ML v1.9 added TFHE-rs ciphertext compatibility so Rust-based TFHE-rs pipelines can exchange encrypted ML inputs and outputs. SE018
CE022 Concrete ML v1.9 says the encrypted LoRA fine-tuning example reaches up to 64 tokens per second on a desktop GPU. SE018
CE023 Concrete ML v1.9 says the full fine-tuning example takes 28 hours across 50 desktop GPUs. SE018
CE024 The protocol announcement says Zama combines FHE with MPC and ZK to achieve confidential yet publicly verifiable smart contracts. SE004
CE025 Zama’s homepage claims the protocol currently processes about 20 transactions per second per chain and targets 1,000 next year plus 10,000+ TPS with ASICs. SE001
CE026 The July 2025 release roundup tied TFHE-rs v1.3, Concrete v2.11, a new Concrete ML demo, and FHEVM v0.7 into a monthly release cadence. SE020
CE027 The public testnet update reports more than 1.2 million encrypted transactions, over 19,000 confidential contracts, 120,000+ active wallets, and 20+ partners building. SE021, SE028
CE028 The same update says testnet threshold key generation and decryption are operated by 13 independent MPC nodes. SE021, SE024
CE029 Zama says the latest protocol update delivered a 6.6x public-decryption improvement and a 19.2x user-decryption improvement versus prior versions. SE021
CE030 The protocol update says Zama dedicated nearly 70 audit-weeks to first-release reviews across TFHE-rs, KMS, coprocessor, gateway, token, and governance contracts. SE021
CE031 The protocol update says official ERC-7984 wrappers are live on mainnet and testnet for USDC, USDT, WETH, BRON, ZAMA, tGBP, and XAUt. SE022
CE032 The protocol update says the beta SDK ships as @zama-fhe/sdk and @zama-fhe/react-sdk and hides protocol complexity behind clear-text developer abstractions. SE022, SE027
CE033 Delegated decryption is presented as a rules-based primitive that lets custodians, compliance providers, or regulators decrypt only specific encrypted values. SE022, SE027
CE034 The protocol update says more than 50 percent of circulating ZAMA had already been staked through operator pools at the time of publication. SE022, SE027
CE035 The mainnet developer program post says the protocol went live on mainnet on December 31, 2025. SE023
CE036 The same post says the first application, Zama Auction, launched on January 21, 2026 on mainnet using sealed-bid FHE. SE023
CE037 The developer program describes builder, bounty, startup, and special payroll tracks as the main community funnels for new applications. SE023
CE038 The status page reported 99.964 percent uptime for MPC mainnet and 100 percent uptime for the mainnet coprocessor at fetch time. SE025
CE039 The same status page showed materially weaker public uptime on at least one testnet service, with the testnet coprocessor at 85.522 percent over the displayed window. SE025
CE040 AWS says Zama uses powerful cloud instances because practical FHE for machine learning and blockchain requires significant computing power. SE026
CE041 AWS highlights Concrete ML for encrypted ML inference and the Zama protocol for private stablecoins and selectively auditable smart-contract flows. SE026
CE042 Grafa reports the live stack includes confidential portfolio management, staking, and bridging across Ethereum, BNB Smart Chain, Hyperliquid, and Solana. SE027
CE043 Bankless describes Zama as wrapping privacy around existing chains instead of forcing builders onto a separate confidentiality chain. SE028
CE044 Bankless identifies Bron, Raycash, Zaiffer, and TokenOps as applications preparing for or using the Zama mainnet stack. SE028
CE045 KuCoin characterizes 2026 FHE as practical enough for production pilots while still carrying a material performance penalty versus plaintext computation. SE029
CE046 KuCoin compares Zama with other privacy stacks such as Fhenix and Inco, indicating that confidential-computation leadership is still contested. SE029
CE047 Omakase describes the stack as a combination of host chain, confidential contracts, coprocessors, gateway, KMS, and relayer layers rather than a single monolithic component. SE030
CE048 Omakase says enterprise deployment realism depends on reusing existing L1 or L2 environments instead of moving users onto a dedicated privacy chain. SE030
CE049 The GitHub repositories and package registries provide public developer-signal for Zama across Rust and Python ecosystems. SE009, SE010, SE011, SE013, SE015
CE050 Zama’s open-source repositories state that commercial use requires a patent license even though development and experimentation are open under BSD-3-Clause-Clear. SE009, SE010, SE011
CU001 The ecosystem page says apps and infrastructure are live on the Zama testnet and or mainnet. SU001
CU002 The ecosystem page lists Bron as a mainnet wallet with native FHE, MPC, biometric passkeys, and social recovery. SU001, SU012
CU003 The ecosystem page lists Raycash as a testnet payments app built around private digital dollars, IBANs, and cards. SU001
CU004 The ecosystem page lists T-REX Network as an RWA product built on the ERC-3643 standard and developed by Tokeny. SU001, SU018
CU005 Zama’s Kinexys post describes a proof of concept inside the Kinexys Digital Assets sandbox focused on private fund subscriptions, blind-pool secondary trading, DvP settlement, and encrypted KYC AML checks. SU002
CU006 The T-REX partnership post says more than 32 billion dollars of assets already live onchain under ERC-3643. SU003, SU017
CU007 The same post says Apex Group committed to placing 100 billion dollars of tokenized assets on the T-REX Ledger by June 2027. SU003, SU019
CU008 Zama positions its protocol as the native confidentiality layer for the T-REX Ledger rather than a peripheral add-on. SU003, SU004
CU009 The GSR official announcement says GSR executed the first confidential OTC trade on Ethereum using the Zama protocol between fully KYCd counterparties. SU006, SU005
CU010 GSR says confidential settlement infrastructure addresses one of the largest pain points for institutions trading digital assets. SU006
CU011 Zama’s Dfns partnership post says Dfns integrated the confidential token standard into its wallet and transaction stack. SU007
CU012 The Dfns partnership says encrypted onchain transactions become available to more than 400 enterprise clients across banking, fintech, custody, RWA, and payments. SU007, SU013
CU013 Zama says Dfns-supported use cases are live now for settlement, B2B payments, tokenized RWA distribution, and compliance through selective disclosure. SU007, SU013
CU014 The Dfns homepage claims the platform delivers over 99.95 percent uptime and highlights reference customers including ABN AMRO, Tokeny, Vita Wallet, Zodia Custody, Sphere, and Nilos. SU008
CU015 Zama’s OpenZeppelin partnership frames the relationship around confidential smart-contract standards and audited Solidity tooling rather than a direct revenue customer logo. SU009
CU016 The Blockscout posts say confidential ERC-7984 token balances and transfer amounts stay encrypted while onchain activity remains publicly verifiable. SU010, SU011
CU017 The payroll mainnet announcement says Bron executed the first confidential payroll on Ethereum mainnet using confidential USDT. SU012
CU018 The same payroll post says Bron is the first wallet to support confidential token features natively. SU012
CU019 The Q1 2026 Shielded report says institutional partners moved into production during the quarter. SU013
CU020 The Shielded report says 51 of 880 projects in the PL Genesis hackathon were built on the Zama protocol. SU013
CU021 The Shielded report says T-REX, OpenZeppelin, Raycash, and Orion Finance publicly presented at Zama Builder Villa around confidential onchain finance use cases. SU013
CU022 The Season 1 winners post says builders shipped encrypted wallets, confidential payroll infrastructure, private legal platforms, and FHE-powered AI payment rails on top of the Zama protocol. SU014
CU023 The same winners post explicitly warns that the listed community projects are unaudited third-party apps. SU014
CU024 Season 2 of the developer program offered more than 15,000 cUSDT across builder, bounty, and APAC tracks for confidential-finance apps and tooling. SU015
CU025 Season 2 says submissions can deploy either on Sepolia testnet or Ethereum mainnet, indicating the funnel spans both experimental and live environments. SU015
CU026 The protocol update says the full protocol stack is live and integrable on mainnet, including wrappers, portfolio, staking, and bridge apps. SU016
CU027 The protocol update says official wrappers currently support USDC, USDT, WETH, BRON, ZAMA, tGBP, and XAUt. SU016
CU028 The ERC-3643 site says the standard is an open-source suite of permissioned-token contracts with built-in identity controls and cites more than 32 billion dollars of tokenized assets. SU017
CU029 Tokeny positions itself as an onchain finance operating system, reinforcing that Zama’s T-REX path is embedded in a larger compliance-first stack. SU018
CU030 Chainwire independently repeats that T-REX is supported by Apex Group, is anchored in ERC-3643, and aims at large-scale institutional RWA tokenization. SU019
CU031 Cointelegraph says Zama’s T-REX integration enters an industry debate where zero-knowledge systems, permissioned networks, and FHE are all competing to become part of the tokenization stack. SU020
CU032 Cointelegraph reports Rand Hindi said FHE can add a few seconds of latency while leaving T-REX throughput and public-chain composability intact. SU020
CU033 The Daily Hodl repeats the T-REX plus Apex commitment and frames the partnership as institutional-grade confidentiality infrastructure for RWA tokenization. SU021
CU034 Bitget independently summarizes the first confidential OTC transaction on Ethereum between Zama and GSR. SU022
CU035 Bankless says more than 120,000 addresses conducted more than 1.2 million encrypted transactions before mainnet broadening, and it spotlights Bron, Raycash, TokenOps, and Zaiffer as app examples. SU023
CU036 KuCoin argues that FHE became practical for pilots in 2026 but still carries a notable performance penalty versus plaintext computation, even after major improvements. SU024
CU037 KuCoin compares Zama with Fhenix and Inco, indicating that public-chain confidentiality is still a contested market rather than a settled category. SU024
CU038 Omakase says the strongest early Zama use cases are confidential tokens, private transfers, auctions, identity-linked compliance flows, and credit or eligibility logic. SU025
CU039 The KuCoin flash item independently repeats the T-REX confidentiality integration for institutional RWA tokenization. SU026
CU040 The customer proof base is strongest in institutional finance infrastructure, wallets, tokenization rails, and ecosystem tooling rather than in disclosed broad enterprise-seat deployments. SU002, SU003, SU006, SU007, SU012, SU016
CU041 Public evidence for customer durability is mostly proxy-based because no fetched source disclosed NRR, GRR, churn, or top-customer concentration. SU007, SU013, SU016
CU042 Dfns, T-REX, and GSR together imply an expansion path from infrastructure integration into settlement, payments, custody, tokenized assets, and OTC trading workflows. SU006, SU007, SU008, SU019
CU043 The visible flagship logos are partner-anchored enough that channel dependence and concentration risk remain real diligence issues. SU003, SU006, SU007, SU015, SU020
CU044 Because Zama itself labels many ecosystem projects as unaudited and because several programs still route through testnet, ecosystem breadth should not be mistaken for underwritten revenue quality. SU014, SU015
CR001 Zama's public website is operated by Zama SAS in Paris and identifies Rand Hindi as CEO. SR002, SR003
CR002 Zama's website terms submit website disputes to French law and the courts under the Paris Court of Appeal. SR003, SR005
CR003 Zama's corporate website privacy policy was last updated on 2023-03-01 and is framed around GDPR plus French data-protection law. SR004
CR004 Zama Switzerland AG, not Zama SAS, is the issuer named in the token sale and token auction legal terms. SR006, SR007, SR008
CR005 The token sale privacy notice says Zama Switzerland AG processes personal data for auction.zama.org and uses Sumsub-supplied ID and proof-of-address information for AML purposes. SR008
CR006 The token auction terms require participants not to be in comprehensively sanctioned jurisdictions and not to appear on Swiss, EU, U.S., or other sanctions lists. SR006, SR007
CR007 The token sale terms also exclude participants from jurisdictions designated as non-cooperative with international AML principles or procedures. SR006, SR007
CR008 The staking terms say the interface is functionally and legally distinct from the protocol and that use of the interface is undertaken at the user's own risk. SR009
CR009 ESMA describes MiCA as instituting uniform EU market rules for crypto-assets, including transparency, disclosure, authorization, and supervision obligations. SR021
CR010 ESMA states MiCA entered into force in June 2023 and required sequential Level 2 and Level 3 implementation measures before the regime fully applied. SR021
CR011 EU Regulation 2023/1113 requires information on originators and beneficiaries to accompany certain crypto-asset transfers to support transfer traceability. SR022, SR023
CR012 The FATF says its virtual-asset standards are designed to make virtual-asset and VASP activity traceable across jurisdictions. SR023, SR022
CR013 FinCEN guidance confirms that certain virtual-currency business models can fall within U.S. money-services and AML regulation. SR024
CR014 CNIL's GDPR developer guidance emphasizes privacy-by-design and data minimization duties that are directionally stricter than a pure growth-first token onboarding motion. SR025, SR004
CR015 Zama markets confidential tokens as compatible with compliance because designated auditors or regulators can be granted decryption rights. SR012, SR018, SR027
CR016 Zama's confidential-onchain-finance page says ERC-7984 is designed so balances and amounts remain encrypted while access control can be customized to issuer compliance requirements. SR010, SR012
CR017 The July 2025 public testnet update reports more than 1.2 million encrypted transactions, more than 19,000 confidential contracts, more than 120,000 active wallets, and 20+ partners. SR011
CR018 The same testnet update calls FHEVM v0.9 the first mainnet release candidate and reports a 10x improvement in decryption performance. SR011
CR019 Zama said threshold key generation and decryption on the updated testnet are handled by a network of 13 MPC nodes. SR011
CR020 Zama's mainnet DKG ran from 2025-11-25 to 2025-11-28 with 13 independent parties and no crashes, dropouts, or invalid messages. SR014
CR021 The DKG required 10.5 TB of data sent per party, showing that Zama's privacy stack still carries heavyweight operational complexity. SR014
CR022 Mainnet staking launched with 18 operators split across 5 FHE nodes and 13 KMS nodes. SR013
CR023 Zama's initial staking emissions are set at 5% yearly and rewards are split 40% to FHE nodes and 60% to KMS nodes. SR013
CR024 The list of operators includes external firms such as Artifact, Blockscape, Conduit, DFNS, Etherscan, Figment, Fireblocks, InfStones, LayerZero, Ledger, Omakase, OpenZeppelin, P2P, Stake Capital, Unit 410, and Zama itself. SR011, SR013
CR025 Zama's post-launch update says the protocol stack is live and integrable but the SDK release is still labeled beta. SR012
CR026 The public auction report says Zama took only three days to push total value shielded above $100 million without downtime. SR006, SR013
CR027 The same auction report says $118.5 million of value was committed but only $44 million was ultimately paid by winning bids, implying material oversubscription and refund-driven demand volatility. SR006, SR013
CR028 The auction report lists 24,697 total bids, 11,103 unique bidders, a $0.05 clearing price, and 218% oversubscription. SR006, SR013
CR029 The staking launch post says auction participants could start staking immediately once claiming opened on 2026-02-02, tightly coupling token distribution to security participation. SR013
CR030 J.P. Morgan's Project EPIC says tokenized finance needs privacy, identity, and composability, but it frames the work as a paper and proof-of-concept rather than scaled production. SR028, SR017
CR031 Zama's T-REX Ledger announcement says more than $32 billion of assets already live onchain under ERC-3643 and that Apex Group committed to place $100 billion on the T-REX Ledger by June 2027. SR016
CR032 The Dfns partnership announcement says encrypted onchain transactions are available to 400+ enterprise clients across banking, fintech, custody, RWA, and payments. SR018
CR033 The GSR announcement describes the Ethereum OTC trade as a proof-of-concept, which is helpful validation but still short of repeatable production volume. SR019
CR034 The Bron payroll announcement says a confidential payroll was executed on Ethereum mainnet using confidential USDT, making payroll one live production use case. SR015
CR035 The privacy-coin regulatory sources argue that stronger KYC, AML, and exchange restrictions remain a structural headwind for assets marketed primarily around privacy. SR032, SR033
CR036 Sidley's 2026 blockchain outlook says compliant confidential-blockchain and tokenization architectures face increasing legal and regulatory complexity. SR026
CR037 Zama's about page says the company has 96 people and 37 PhDs, indicating a research-heavy team relative to the breadth of commercialization, compliance, and operator-management work now underway. SR002
CR038 Zama says it licensed its FHE technology to dozens of companies across blockchain and AI before launching the protocol. SR001
CR039 Funding coverage from Zama, The Block, EU-Startups, and Financial IT consistently describes the Series B as $57 million at a valuation above $1 billion. SR001, SR029, SR030, SR031
CR040 The combination of beta developer tooling, heavyweight cryptographic ceremonies, and third-party operator dependence means residual operational risk remains high even after mainnet launch. SR011, SR012, SR013, SR014
CR041 If MiCA, travel-rule, sanctions, or KYC controls fail in practice, Zama's pitch to regulated finance customers can break before revenue scales. SR006, SR007, SR021, SR022, SR023, SR024
CR042 Public sources still do not disclose Zama's revenue, conversion from pilots to production contracts, or audited uptime metrics, leaving investors unable to verify commercialization quality.
CV001 Zama's June 2025 Series B was publicly described as $57 million at a valuation above $1 billion. SV001, SV012, SV013, SV014
CV002 Alpha Drops lists a Jan. 26, 2026 public token sale that raised $44 million at a $550 million valuation. SV015
CV003 Premier Alternatives lists a market-implied valuation of $840.3 million for Zama and total funding of $175.6 million. SV016
CV004 Zama's auction announcement said 12% of total token supply would be distributed through public sales including a sealed-bid Dutch auction. SV003
CV005 Zama's auction-demand post reports $118.5 million committed, $44 million paid by winning bids, a $0.05 clearing price, and 218% oversubscription. SV004, SV015
CV006 The same auction-demand post reports 2.805 billion tokens demanded versus 880 million tokens sold. SV004
CV007 Bitget's Zama report says the protocol has an 11 billion token supply and that the January 2026 auction app briefly became the highest-volume application on Ethereum. SV017, SV004
CV008 Zama says $ZAMA has two main utilities today: paying encryption or decryption fees that are burned and staking to secure operators. SV005
CV009 Zama's staking launch set initial yearly emissions at 5% and disclosed 18 operators split between FHE and KMS roles. SV005
CV010 Zama's shielded update says more than 50% of circulating supply is staked across 18 active operators. SV007, SV005
CV011 Zama's post-launch update says the full protocol stack is live and integrable since mainnet launch, but the SDK is still in beta. SV006
CV012 The Bron payroll announcement says a confidential payroll was executed on Ethereum mainnet using confidential USDT. SV008
CV013 The DKG announcement says 13 parties completed mainnet key generation without crashes, dropouts, or invalid messages. SV009
CV014 The shielded report says Dfns brings encrypted transactions to 400+ enterprise clients. SV007
CV015 J.P. Morgan's Project EPIC says tokenized finance needs privacy, identity, and composability to scale. SV034, SV018
CV016 Ripple and BCG say tokenized real-world assets could exceed $18 trillion by 2033, roughly 30x larger than today. SV019, SV020
CV017 Research and Markets says the tokenization market is worth $5.19 billion in 2026 and could reach $13.2 billion by 2030 at a 26.3% CAGR. SV021, SV022
CV018 Zama's about page says the company has 96 people and 37 PhDs. SV002
CV019 Premier Alternatives values Fireblocks at $8 billion and says it has raised $1 billion. SV026
CV020 Premier Alternatives values Digital Asset at $380 million and says it has raised $457.2 million. SV027
CV021 Premier Alternatives shows StarkWare has raised $261.5 million, but the current valuation is not displayed publicly on the fetched page. SV028, SV030
CV022 Yahoo Finance showed Arqit at about $280.35 million market capitalization and 233.58x enterprise value to revenue on May 27, 2026. SV023
CV023 Arqit reported first-half fiscal 2026 revenue of $623,000 and cash of $28.9 million as of March 31, 2026. SV025, SV024
CV024 Arqit demonstrates that listed cryptography infrastructure can show optically huge revenue multiples while still operating at very small absolute revenue scale. SV023, SV025
CV025 Fireblocks is the strongest private multi-billion comparison in this set because it combines institutional crypto infrastructure with much broader distribution than Zama currently discloses. SV026, SV029
CV026 Digital Asset shows that enterprise ledger infrastructure can trade far below Zama's last private round despite substantial cumulative funding. SV027, SV016
CV027 Zama's public evidence does not disclose revenue, ARR, gross margin, or customer retention.
CV028 Because revenue is undisclosed, public valuation support depends more on market position, adoption signals, and capital-market datapoints than on cash-flow multiples. SV001, SV015, SV016, SV019
CV029 The public equity round (> $1 billion), secondary-style mark ($840.3 million), and token-sale implied valuation ($550 million) create a wide pricing band rather than a single settled value. SV001, SV015, SV016
CV030 The token-sale implied valuation is not directly equivalent to equity value, but it is still the clearest live market-clearing datapoint for protocol demand that is public today. SV004, SV015
CV031 CoinLaw and KuCoin both describe persistent compliance and exchange headwinds for privacy-linked crypto assets in 2026. SV031, SV032
CV032 Sidley's 2026 blockchain outlook says legal and regulatory complexity for tokenized and confidential architectures remains high. SV033
CV033 Figment argues Zama is building programmable confidentiality with compliance hooks rather than the fully opaque posture associated with mixers or legacy privacy coins. SV018, SV010
CV034 Zama says it had already licensed technology to dozens of companies across blockchain and AI before unveiling the protocol. SV001
CV035 Zama's headcount remains below 100 while public valuation markers range from $550 million to above $1 billion, implying demanding execution expectations per employee. SV002, SV015, SV016
CV036 If mainnet fee burn and enterprise transaction volume do not ramp, emission-based staking economics alone are unlikely to justify premium pricing. SV005, SV007, SV027
CV037 A credible bull case requires Dfns-style channel distribution, repeat enterprise workflows, and evidence that Zama converts production use cases into sustained protocol fees. SV007, SV008, SV010, SV034
CV038 A credible bear case is that regulation and slow commercialization compress value toward the token-sale or secondary-market range instead of the >$1 billion equity mark. SV015, SV016, SV031, SV032, SV033
CV039 Public evidence supports a price-sensitive research-more stance rather than a buy call because valuation support is still evidence-sensitive, not purely narrative-driven. SV015, SV016, SV027, SV033
CV040 A fair public-evidence entry range is closer to roughly $550 million to $840 million than above $1 billion because disclosed metrics stop short of revenue proof. SV015, SV016, SV027
CV041 The best current recommendation is research-more with medium confidence, high risk, and a stretched stance above the last disclosed >$1 billion equity round. SV015, SV016, SV027, SV033
CV042 The next diligence unlock is audited revenue, gross-margin, customer-concentration, and cap-table disclosure rather than another top-down market-size slide.
来源
编号出版方标题引文
SO001 Zama About Zama - Pioneering Fully Homomorphic Encryption And State-Of-The-Art Cryptography Zama is an open source cryptography company building state-of-the-art Fully Homomorphic Encryption (FHE) solutions for blockchain.
SO002 Zama Zama - Open Source Cryptography
SO003 Zama Legal Notice - Zama The Website is published, owned and operated by Zama SAS ... whose registered office is located at 8 rue du Sentier, Paris 75002.
SO004 INPI Entreprise : ZAMA SAS - SIREN 879 243 319
SO005 Amazon Web Services Customer Testimonial: Zama
SO006 Zama The Zama FHE Master Plan
SO007 Zama Announcing Our Series B and the Zama Confidential Blockchain Protocol
SO008 Zama Zama becomes the confidentiality layer for the T-REX Ledger
SO009 Zama The confidentiality layer: Bringing institutional trade execution to Ethereum with GSR
SO010 Zama Zama partners with OpenZeppelin to bring confidential smart contracts to DeFi and digital assets
SO011 Zama Kinexys by JPMorgan releases a proof-of-concept leveraging Zama FHEVM
SO012 Zama Dfns and Zama partner to unlock institutional liquidity on public blockchains through confidential wallet infrastructure
SO013 Zama Shielded report, Dfns partnership and new Zama SDK
SO014 Zama Protocol Docs FHE on blockchain | Zama Protocol
SO015 Zama Protocol Docs Zama Confidential Blockchain Protocol Litepaper
SO016 GitHub Zama · GitHub
SO017 GitHub zama-ai/tfhe-rs
SO018 GitHub zama-ai/concrete-ml
SO019 GitHub zama-ai/fhevm
SO020 TechCrunch Zama’s homomorphic encryption tech lands it $73M on a valuation of nearly $400M Homomorphic encryption ... continues to elude mass-market scalability and thus adoption.
SO021 Tech.eu Zama becomes 1st fully homomorphic encryption unicorn with $57M raise
SO022 EU-Startups Paris-based cryptography startup Zama secures €49 million to bring end-to-end encryption to public blockchains
SO023 Verdict Zama secures $57m for blockchain encryption technology
SO024 SiliconANGLE Zama raises $57M to build full encryption for public blockchains
SO025 Chainwire T-REX Network and Zama launch institutional-grade confidentiality infrastructure for RWA tokenization
SO026 Cointelegraph Cryptography firm Zama taps T-REX to bring bank-grade privacy to public blockchains
SO027 Dfns We’re joining the Zama Network
SO028 CoinGabbar Why Is Zama Price Crashing Today? Will Coinstore Listing Fuel Recovery The token debuted on February 2 at $0.03837 ... Since then, value has dropped nearly 50%.
SO029 Phemex Zama Protocol Launches Mainnet on Ethereum, Plans Token Listing in April 2026
SO030 MEXC Zama Analysis: Is A $1 Billion USD Valuation Justified? This has sparked growing interest and debate around whether such a valuation is truly justified.
SM001 Zama Protocol Docs Zama Confidential Blockchain Protocol Litepaper
SM002 Zama Protocol Docs FHE on blockchain | Zama Protocol
SM003 Figment Inside Zama: How Fully Homomorphic Encryption Brings Programmable Privacy to Public Blockchains
SM004 Chainwire T-REX Network and Zama launch institutional-grade confidentiality infrastructure for RWA tokenization
SM005 Cointelegraph Cryptography firm Zama taps T-REX to bring bank-grade privacy to public blockchains
SM006 Dfns We’re joining the Zama Network
SM007 Zama Zama partners with OpenZeppelin to bring confidential smart contracts to DeFi and digital assets
SM008 Zama Kinexys by JPMorgan releases a proof-of-concept leveraging Zama FHEVM
SM009 Zama Dfns and Zama partner to unlock institutional liquidity on public blockchains through confidential wallet infrastructure
SM010 Zama Announcing Our Series B and the Zama Confidential Blockchain Protocol
SM011 Forbes Stablecoin Volume Hits Record $4.5T As Asia Drives Q1 Boom
SM012 a16z crypto 6 trends for 2026: Stablecoins, payments, and real-world assets
SM013 IMF Stablecoins and the Future of Payments: Evidence from Financial Markets
SM014 Ledger Insights McKinsey estimates tokenization will be less than $2 trillion by 2030
SM015 Boston Consulting Group Tokenized Funds: The Third Revolution in Asset Management Decoded
SM016 Crowdfund Insider Tokenized Real-World Assets (RWAs) Surge to $19.3B by Q1 2026 as Institutional Momentum Builds
SM017 RWA.xyz RWA.xyz market dashboard
SM018 AssetTokenization.com Asset Tokenization Forecasts Range: $2T to $30T by 2030
SM019 Phemex Zama Protocol Launches Mainnet on Ethereum, Plans Token Listing in April 2026
SM020 Phemex ZAMA Price Prediction 2026-2031: Post-Launch Analysis & Phemex Guide
SM021 MEXC Zama Analysis: Is A $1 Billion USD Valuation Justified?
SM022 CoinGabbar Why Is Zama Price Crashing Today? Will Coinstore Listing Fuel Recovery
SM023 Blockchain.News ZAMA Collapse to $0.025 Within 72 Hours as Technical Structure Implodes
SM024 Verdict Zama secures $57m for blockchain encryption technology
SM025 EU-Startups Paris-based cryptography startup Zama secures €49 million to bring end-to-end encryption to public blockchains
SM026 Swift Unlocking the potential of asset tokenisation
SP001 Zama Announcing Our Series B and the Zama Confidential Blockchain Protocol
SP002 Zama About Zama - Pioneering Fully Homomorphic Encryption And State-Of-The-Art Cryptography A team of world-renowned cryptographers, engineers and entrepreneurs solving some of the biggest privacy challenges.
SP003 Zama Kinexys by J.P. Morgan Releases a Proof-of-concept Focused on the Financial Sector Leveraging Zama’s Privacy-Preserving Solutions
SP004 Zama Dfns and Zama Partner to Unlock Institutional Liquidity on Public Blockchains Through Confidential Wallet Infrastructure
SP005 Fhenix Fhenix | Ethereum's privacy with Fully Homomorphic Encryption (FHE)
SP006 Fhenix Docs Introduction - Fhenix Documentation
SP007 Inco Inco — Full-Stack Privacy for Blockchains
SP008 Secret Network Docs Secret Network Introduction | Secret Network
SP009 Secret Network Secret Network
SP010 Oasis Docs Sapphire ParaTime | Oasis Documentation
SP011 Oasis Docs Getting Started | Oasis Documentation
SP012 Zama Zama Becomes the Confidentiality Layer for the T-REX Ledger
SP013 Aztec Docs Aztec Documentation | Privacy-first zkRollup | Aztec Documentation
SP014 Aztec The first decentralized, privacy-preserving L2 on Ethereum.
SP015 Aleo Docs Aleo Documentation | Aleo Docs
SP016 Aleo Announcing the Aleo Bug Bounty Program
SP017 Partisia Docs Partisia Platform Documentation
SP018 Amazon Web Services What is Nitro Enclaves? - AWS Nitro Enclaves
SP019 Microsoft Azure Azure confidential computing
SP020 Google Cloud Confidential Space overview  |  Google Cloud Documentation
SP021 Arcium Docs Overview of the Arcium Network - Arcium Docs
SP022 Figment Inside Zama: How Fully Homomorphic Encryption Brings Programmable Privacy to Public Blockchains
SP023 ChainSafe The 2026 Guide to Blockchain Privacy: 5 approaches for institutional adoption
SP024 KuCoin FHE 2026: Private Blockchains & Encrypted Data | KuCoin
SP025 BlockEden Navigating the Privacy Technology Landscape: FHE, ZK, and TEE in Blockchain - BlockEden.xyz
SI001 Zama Announcing Our Series B and the Zama Confidential Blockchain Protocol
SI002 Zama About Zama - Pioneering Fully Homomorphic Encryption And State-Of-The-Art Cryptography
SI003 Zama The Zama Public Auction: $118M Committed for the First Encrypted ICO on Ethereum
SI004 Zama Switzerland AG Token Sale Terms
SI005 Zama Dfns and Zama Partner to Unlock Institutional Liquidity on Public Blockchains Through Confidential Wallet Infrastructure
SI006 Zama Kinexys by J.P. Morgan Releases a Proof-of-concept Focused on the Financial Sector Leveraging Zama’s Privacy-Preserving Solutions
SI007 Zama Zama Becomes the Confidentiality Layer for the T-REX Ledger
SI008 Zama Docs Welcome to Zama's documentation | Homepage
SI009 Zama Docs Welcome | Concrete ML
SI010 Zama Docs Welcome | Concrete
SI011 Zama Docs Welcome to TFHE-rs | TFHE-rs
SI012 Annuaire des Entreprises Zama SAS - Annuaire des Entreprises
SI013 Data INPI ZAMA SAS (Entreprises) - Data INPI
SI014 Pappers Société ZAMA SAS : Chiffre d'affaires, statuts, extrait d'immatriculation
SI015 Societe.com Société ZAMA SAS (879243319) à PARIS (75002).
SI016 TechCrunch Zama's homomorphic encryption tech lands it $73M on a valuation of nearly $400M | TechCrunch
SI017 Tech.eu Zama becomes 1st fully homomorphic encryption unicorn with $57M raise
SI018 The Block Pantera, Blockchange lead funding for privacy-focused Zama Confidential Blockchain Protocol at $1 billion valuation
SI019 CoinList CoinList Launches First Fully Onchain Token Sale with Zama
SI020 Figment Inside Zama: How Fully Homomorphic Encryption Brings Programmable Privacy to Public Blockchains
SI021 Cryptomaniaks Zama Token Launch: HTTPS for Blockchain or Just Hype?
SI022 MEXC Zama's Public Token Auction Approaches: The Shift From 'Privacy Coins' To 'Confidential Computation' Infrastructure
SI023 EU-Startups Paris-based cryptography startup Zama secures €49 million to bring end-to-end encryption to public blockchains | EU-Startups
SI024 Financial IT Zama Raises $57M In Series B To Bring End-to-End Encryption To Public Blockchains
SI025 ICO Analytics Zama (ZAMA) token sale analytics and information, private/seed sale price, tokenomics | ICO Analytics
SE001 Zama Zama - Open Source Cryptography
SE002 Zama About Zama - Pioneering Fully Homomorphic Encryption And State-Of-The-Art Cryptography
SE003 Zama Confidential Onchain Finance
SE004 Zama Announcing Our Series B and the Zama Confidential Blockchain Protocol
SE005 Zama Docs Welcome to Zama's documentation | Homepage
SE006 Zama Docs Welcome | Zama Protocol
SE007 Zama Docs Welcome to TFHE-rs | TFHE-rs
SE008 Zama Docs Welcome | Concrete ML
SE009 GitHub GitHub - zama-ai/tfhe-rs: TFHE-rs: A Pure Rust implementation of the TFHE Scheme for Boolean and Integer Arithmetics Over Encrypted Data.
SE010 GitHub GitHub - zama-ai/concrete-ml: Concrete ML: Privacy Preserving ML framework using Fully Homomorphic Encryption (FHE), built on top of Concrete, with bindings to traditional ML frameworks.
SE011 GitHub GitHub - zama-ai/fhevm: FHEVM, a full-stack framework for integrating Fully Homomorphic Encryption (FHE) with blockchain applications
SE012 GitHub GitHub - zama-ai/fhevm-solidity: A Solidity library for interacting with fhevm.
SE013 crates.io crates.io: Rust Package Registry
SE014 docs.rs tfhe - Rust
SE015 PyPI concrete-ml
SE016 PyPI concrete-python
SE017 Zama TFHE-rs v1.4: GPU Performance Breakthrough and More
SE018 Zama Concrete ML v1.9: TFHE-rs Compatibility and Faster LLM Fine-tuning
SE019 Zama fhEVM v0.6: Enhanced Input Mechanism with Proof Capabilities, Expanded type, and Better Configurability
SE020 Zama Zama Product Releases - July 2025
SE021 Zama Zama Protocol Testnet Update: MPC Partners, Better Performance, Audits and New Features
SE022 Zama Zama Protocol Update: New SDK, Delegated Decryption, Developer Tooling, and Protocol Apps
SE023 Zama Zama Developer Program Mainnet Season 1: Building For The Long Game
SE024 Zama Decentralized Key Generation for Zama Mainnet Has Successfully Completed
SE025 Zama Status zama status
SE026 Amazon Web Services Customer Testimonial: Zama
SE027 Grafa Zama launches full confidential blockchain protocol stack
SE028 Bankless ‘Confidentiality Layer’ Zama Wraps Blockchains in Privacy on Bankless
SE029 KuCoin FHE 2026: Private Blockchains & Encrypted Data | KuCoin
SE030 Omakase Zama, FHE, and the Emerging Privacy Stack for Blockchains | Omakase
SU001 Zama Ecosystem
SU002 Zama Kinexys by J.P. Morgan Releases a Proof-of-concept Focused on the Financial Sector Leveraging Zama’s Privacy-Preserving Solutions
SU003 Zama Zama Becomes the Confidentiality Layer for the T-REX Ledger
SU004 Zama Becoming the confidentiality layer for the T-REX Ledger, executing the first confidential OTC trade onchain, presenting at FHE.org 2026, and more
SU005 Zama The Confidentiality Layer: Bringing Institutional Trade Execution to Ethereum with GSR
SU006 GSR GSR and Zama Complete Landmark First Confidential OTC Trade on…
SU007 Zama Dfns and Zama Partner to Unlock Institutional Liquidity on Public Blockchains Through Confidential Wallet Infrastructure
SU008 Dfns Dfns — The Most Powerful Digital Asset Wallet Infrastructure
SU009 Zama Zama Partners with OpenZeppelin to Bring Confidential Smart Contracts to DeFi and Digital Assets
SU010 Zama Blockscout Now Natively Supports ERC-7984 Confidential Tokens
SU011 Blockscout Exploring Confidential Tokens with Zama on Blockscout
SU012 Zama The End of the Open Book: Zama and Bron Execute the First Confidential Payroll on Ethereum Mainnet
SU013 Zama Shielded report, Dfns brings encrypted transactions to 400+ enterprise clients, new SDK and more
SU014 Zama Announcing the Developer Program Mainnet Season 1 Winners
SU015 Zama Zama Developer Program Mainnet Season 2: Confidential Finance Is The Next Frontier
SU016 Zama Zama Protocol Update: New SDK, Delegated Decryption, Developer Tooling, and Protocol Apps
SU017 ERC3643 Association ERC3643 - The Token Standard for RWA Tokenization
SU018 Tokeny Tokeny – The Leading Onchain Finance Operating System
SU019 Chainwire T-REX Network and Zama Launch Institutional-Grade Confidentiality Infrastructure for RWA Tokenization - Chainwire
SU020 Cointelegraph Cryptography Firm Zama Brings FHE Privacy to T‑REX Ledger
SU021 The Daily Hodl T-REX Network and Zama Launch Institutional-Grade Confidentiality Infrastructure for RWA Tokenization - The Daily Hodl
SU022 Bitget Zama and GSR complete the first confidential OTC trade on Ethereum | Bitget News
SU023 Bankless ‘Confidentiality Layer’ Zama Wraps Blockchains in Privacy on Bankless
SU024 KuCoin FHE 2026: Private Blockchains & Encrypted Data | KuCoin
SU025 Omakase Zama, FHE, and the Emerging Privacy Stack for Blockchains | Omakase
SU026 KuCoin T-REX Network and Zama Integrate FHE for Institutional RWA Confidentiality | KuCoin
SU027 FHE.org Fully Homomorphic Encryption
SR001 Zama Announcing Our Series B and the Zama Confidential Blockchain Protocol Today, we are unveiling our most ambitious product to date: the Zama Confidential Blockchain Protocol, and announcing our $57m Series B at a valuation of over $1b.
SR002 Zama About Zama - Pioneering Fully Homomorphic Encryption And State-Of-The-Art Cryptography
SR003 Zama Legal Notice - Zama
SR004 Zama Privacy Policy - Zama
SR005 Zama Terms and conditions - Zama
SR006 Zama Switzerland AG Token Auction Terms
SR007 Zama Switzerland AG Token Sale Terms
SR008 Zama Switzerland AG Token Sale Privacy Notice
SR009 Zama Zama Staking Terms
SR010 Zama Welcome to Zama's documentation | Homepage
SR011 Zama Zama Protocol Testnet Update: MPC Partners, Better Performance, Audits and New Features
SR012 Zama Zama Protocol Update: New SDK, Delegated Decryption, Developer Tooling, and Protocol Apps
SR013 Zama Staking is Now Live on Mainnet
SR014 Zama Decentralized Key Generation for Zama Mainnet Has Successfully Completed
SR015 Zama The End of the Open Book: Zama and Bron Execute the First Confidential Payroll on Ethereum Mainnet
SR016 Zama Zama Becomes the Confidentiality Layer for the T-REX Ledger
SR017 Zama Kinexys by J.P. Morgan Releases a Proof-of-concept Focused on the Financial Sector
SR018 Zama Dfns and Zama Partner to Unlock Institutional Liquidity on Public Blockchains Through Confidential Wallet Infrastructure
SR019 Zama The Confidentiality Layer: Bringing Institutional Trade Execution to Ethereum with GSR
SR020 Zama Zama Partners with OpenZeppelin to Bring Confidential Smart Contracts to DeFi and Digital Assets
SR021 European Securities and Markets Authority Markets in Crypto-Assets Regulation (MiCA)
SR022 European Union Regulation (EU) 2023/1113 on information accompanying transfers of funds and certain crypto-assets
SR023 Financial Action Task Force Virtual Assets
SR024 FinCEN Application of FinCEN's Regulations to Certain Business Models
SR025 CNIL GDPR developer's guide
SR026 Sidley Sidley Blockchain Bulletin - 2026 Business, Legal and Regulatory Outlook
SR027 Figment Zama First Look: Bringing compliant confidentiality on-chain
SR028 J.P. Morgan Kinexys Project Epic | J.P. Morgan
SR029 EU-Startups Paris-based cryptography startup Zama secures €49 million to bring end-to-end encryption to public blockchains
SR030 Financial IT Zama Raises $57M In Series B To Bring End-to-End Encryption To Public Blockchains
SR031 The Block Pantera, Blockchange lead funding for privacy-focused Zama Confidential Blockchain Protocol at $1 billion valuation
SR032 CoinLaw Privacy Coins vs. Regulatory Compliance Statistics
SR033 KuCoin Privacy Coins Gain Momentum in 2026 Amid Regulatory Challenges
SV001 Zama Announcing Our Series B and the Zama Confidential Blockchain Protocol Today, we are unveiling our most ambitious product to date: the Zama Confidential Blockchain Protocol, and announcing our $57m Series B at a valuation of over $1b.
SV002 Zama About Zama - Pioneering Fully Homomorphic Encryption And State-Of-The-Art Cryptography
SV003 Zama Announcing the Zama Public Auction
SV004 Zama The Zama Public Auction: $118M Committed for the First Encrypted ICO on Ethereum
SV005 Zama Staking is Now Live on Mainnet
SV006 Zama Zama Protocol Update: New SDK, Delegated Decryption, Developer Tooling, and Protocol Apps
SV007 Zama Shielded report, Dfns brings encrypted transactions to 400+ enterprise clients, new SDK and more
SV008 Zama The End of the Open Book: Zama and Bron Execute the First Confidential Payroll on Ethereum Mainnet
SV009 Zama Decentralized Key Generation for Zama Mainnet Has Successfully Completed
SV010 Zama Confidential Onchain Finance
SV011 République Française Zama SAS company registry page
SV012 The Block Pantera, Blockchange lead funding for privacy-focused Zama Confidential Blockchain Protocol at $1 billion valuation
SV013 EU-Startups Paris-based cryptography startup Zama secures €49 million to bring end-to-end encryption to public blockchains
SV014 Financial IT Zama Raises $57M In Series B To Bring End-to-End Encryption To Public Blockchains
SV015 Alpha Drops Zama Funding Rounds, Investors & Valuation
SV016 Premier Alternatives Zama - Private Company Valuation & Stock Data
SV017 Bitget Research Report|In-Depth Analysis and Market Cap of Zama (ZAMA) | Bitget News
SV018 Figment Zama First Look: Bringing compliant confidentiality on-chain
SV019 Ripple Tokenization of Financial and Real-World Assets
SV020 Tokenization Policy BCG Tokenization Report Analysis: $16 Trillion Market Projection
SV021 Research and Markets Tokenization Market Report 2026 - Research and Markets
SV022 The Business Research Company Global Tokenization Market Report 2026
SV023 Yahoo Finance Arqit Quantum Inc. (ARQQ) Stock Price, News, Quote & History
SV024 MarketBeat Arqit Quantum (ARQQW) 10K Form and Latest SEC Filings 2026
SV025 Markets Insider Arqit Quantum Inc. Announces Financial Results for First Half of Fiscal Year 2026
SV026 Premier Alternatives Fireblocks - Private Company Valuation & Stock Data
SV027 Premier Alternatives Digital Asset - Private Company Valuation & Stock Data
SV028 Premier Alternatives StarkWare - Private Company Valuation & Stock Data
SV029 PitchBook Fireblocks 2026 Company Profile: Valuation, Funding & Investors | PitchBook
SV030 PitchBook StarkWare 2026 Company Profile: Valuation, Funding & Investors | PitchBook
SV031 CoinLaw Privacy Coins vs. Regulatory Compliance Statistics
SV032 KuCoin Privacy Coins Gain Momentum in 2026 Amid Regulatory Challenges
SV033 Sidley Sidley Blockchain Bulletin - 2026 Business, Legal and Regulatory Outlook
SV034 J.P. Morgan Kinexys Project Epic | J.P. Morgan