LangChain, the company created in response to the explosion in popularity of the original LangChain Python package and that now stands behind a full platform for building 'agents', recently announced it raised $125 million at a $1.25 billion valuation. In addition to providing LangChain with funds, which are always welcome, the Series B officially grants LangChain unicorn status. LangChain's Series B round was led by IVP and existing investors Sequoia, Benchmark, and Amplify. Newcomers CapitalG and Sapphire Ventures also contributed to the round.

The original LangChain package was developer Harrison Chase's side project. Chase built the package as a way to help developers get started on building with LLMs. The LangChain package enabled developers to tackle challenges in multistep workflows, including searching the web and connecting models with external data sources, even before the impactful arrival of ChatGPT. Since the launch of the original package, many other open source and proprietary competitors have emerged. Still, the LangChain package remains a favorite model-neutral option for building complex workflows with LLMs.

Moreover, thanks to its specialized tools, LangGraph and LangSmith, the company has evolved into a full-blown platform for building, customizing and monitoring agent workflows. In fact, alongside its Series B announcement, LangChain detailed it would launch all new versions of the LangGraph and LangChain packages, a new agent in LangSmith Observability, called Insights Agent, which categorizes agent behavior, and the private preview of Agent Builder, a no-code text-based agent building experience.