LangChain is not selling a library first anymore. The homepage puts LangSmith at the center and calls it the agent engineering platform, then breaks the product into observability, evaluation, deployment, and fleet. The open-source side is framed as three paths, deepagents, langchain, and langgraph, which tells you exactly where the team thinks the center of gravity lives now.
The release stream backs that up. ToolVitals shows 30 GitHub releases in 90 days, 27 release events in the last 30 days, and 134,637 GitHub stars. Recent tags keep landing across the stack, including langchain-openai==1.1.15, langchain-anthropic==1.4.1, langchain-core==1.3.0, and langchain-classic==1.0.4. This is not one monolith moving as a single block. It is a package train.
The interesting part is what changed. langchain-core==1.3.0 adds traceable metadata, keeps streaming behavior compatible, and hardens SSRF policy handling. langchain-anthropic==1.4.1 adds Opus 4.7 support and adaptive thinking mode. langchain-openai==1.1.15 fixes streaming dict responses and Azure chat profile inference. The team is still spending cycles on the boring production edges, which is where agent stacks usually break.
What the data does not tell you
A maxed-out ToolVitals score does not prove the product works for your use case. It does not measure code quality, benchmark performance, user satisfaction, revenue, or whether the agents actually behave well under load. It only shows that the project is active, visible, and shipping hard.
How it compares
Among the nearby tools in the data, OpenClaw has more stars, 362,725 versus LangChain’s 134,637, and almost the same release tempo, 28 release events in 30 days versus 27. Continue is slower at 13 release events in 30 days. That puts LangChain in the active maintenance bucket, not the sleepy one.
Bottom line
If your team is building agent workflows and wants an ecosystem with first-party observability, evaluation, and deployment, evaluate LangChain. If you only want a thin wrapper around model calls, this is more platform than you need.
Sources
- https://langchain.com
- https://github.com/langchain-ai/langchain
- https://github.com/langchain-ai/langchain/releases/tag/langchain-core%3D%3D1.3.0
- https://github.com/langchain-ai/langchain/releases/tag/langchain-openai%3D%3D1.1.15
- https://github.com/langchain-ai/langchain/releases/tag/langchain-anthropic%3D%3D1.4.1
- https://github.com/langchain-ai/langchain/releases/tag/langchain-classic%3D%3D1.0.4