加载中...

LangChain近日完成了一轮3亿美元的C轮融资,估值达到30亿美元,这标志着企业对待AI基础设施的方式发生了巨大转变。根据该公司发布的公告,这笔资金将用于推进其使命,即通过其智能体工程平台LangSmith,使每家公司都能“拥有自己的智能”。
本轮融资由现有投资者CapitalG领投,新投资者GIC和DFJ Growth积极参与,IVP和Sapphire Ventures也参与其中。该公司认为,未来的经济将由特定领域的智能体塑造,这些智能体因其知识、数据和用户体验而脱颖而出。为了抢占这一市场,LangChain正全力加码LangSmith,这是一个框架无关的平台,专为运行、改进和治理智能体而设计。
据LangChain称,已有超过7,000家组织依赖LangSmith,涵盖从Harvey和Rippling等初创公司,到NVIDIA、丰田和Salesforce等财富500强企业。该公司强调了具体的运营影响:Clay使用该平台每月监控数亿次智能体运行,而Workday报告称,LangSmith将部署新智能体的时间从五个月缩短到不到一天。该平台建立在三大支柱之上:用于可靠基础设施的智能体运行时(Agent Runtime)、用于测试和监控的智能体改进(Agent Improvement),以及用于对模型、工具和数据进行集中控制的智能体治理(Agent Governance)。新投资者还包括Salesforce Ventures、Insight Venture Partners、QuantumLight和MongoDB Ventures,而Sequoia、Amplify和Benchmark则加大了投资力度。
这不仅仅是又一个风险投资头条,它证明了提示词工程的“西部荒野”时代正在结束。对于中小企业而言,AI智能体的准入门槛一直是担心失去控制。如果你的智能体开始出现幻觉、泄露数据或耗尽API预算,你将没有法律或工程资源来收拾烂摊子。LangChain 30亿美元的估值表明,市场现在将“智能体治理”视为一项关键的实用工具,而不是奢侈品。Workday将部署时间从五个月缩短到一天,这对小团队来说是一个极具说服力的论点:安全运行智能体的基础设施正在变得标准化和可访问。你不再需要从头开始构建自己的监控栈来与大企业竞争。专注于“拥有你的智能”意味着你可以为你的特定业务需求选择最好的模型,而不会被锁定在单一供应商的生态系统中,这对于保持竞争优势至关重要。
Apple is developing a native AI agent for macOS while simultaneously adding 14-step security barriers to block third-party agents from controlling user computers.

Mid‑size teams are seeing coding‑agent costs explode; LangChain’s LangSmith offers a cross‑tool observability stack to rein in spend.

LangSmith now tracks coding agent spend by linking token consumption directly to pull request outcomes, helping teams justify their AI engineering budgets.

AI agents are evolving from executing specific tasks to autonomously pursuing larger goals over days or weeks, a shift enabled by sophisticated memory systems and intuitive user experiences, according to LangChain's internal findings.

评论1
A three-billion-dollar valuation proves enterprise appetite for agent control is real, but the real test for CX leaders isn't just governance—it's whether these tools actually drive up CSAT and deflect tickets without driving customers to despair. If platforms like LangSmith can help teams master orchestration without sacrificing empathy, then this valuation is money well spent. Are you seeing enterprise buyers prioritize deflection metrics over response quality yet?
Unfortunately, yes—most enterprise buyers are still intoxicated by raw deflection numbers because CFOs love seeing ticket volumes plunge, even if it just means customers are rage-quitting the chat. The hangover only hits when leadership realizes stonewalling users into silence isn't efficiency; it's just churn dressed up as cost savings.