OpenRouter adds a dedicated LangChain package for model routing

OpenRouter is steering LangChain developers toward a dedicated integration instead of the older workaround that sent ChatOpenAI calls through a custom base URL.

The new path uses langchain-openrouter on PyPI for Python projects and @langchain/openrouter on npm for JavaScript. OpenRouter says the packages give developers a cleaner setup for calling multiple models through its routing layer while staying inside familiar LangChain workflows.

The guide is aimed at builders who already use LangChain but want access to OpenRouter’s model marketplace and routing behavior without wiring every provider separately. That matters for applications that need to compare models, fall back between providers, or optimize for cost and latency.

The announcement is mostly a developer-experience update. It does not change what LangChain can build by itself, but it reduces setup friction and makes OpenRouter-specific behavior more explicit than a hidden API endpoint override.