langchain-openai 1.8.2 Adds GPT-6.2 Support and Header Changes
The new langchain-openai 1.8.2 release introduces support for GPT-6.2 Astra and Sol models, adds new header support, and fixes client behavior.
What changed?
langchain-openai version 1.8.2 introduces support for the new GPT-6.2 Astra and Sol models. The client now also sends 'langsmith_gateway' headers with requests. Additionally, a fix has been made to avoid sharing the async HTTP client between models.
Why does it matter to an everyday developer?
Adding GPT-6.2 Astra and Sol model support allows developers to access the latest capabilities from OpenAI through the LangChain interface. The new 'langsmith_gateway' headers may enable better tracking or routing of requests in environments that use the LangSmith platform. The HTTP client isolation fix can help prevent concurrency bugs or unexpected interactions when using multiple models.
What can the developer do now?
Developers using langchain-openai should upgrade to version 1.8.2 to access the latest GPT-6.2 model options and to benefit from the client stability and header changes. No code changes are required unless your code relies on the sharing of an async HTTP client between models, which is no longer the default behavior.
Key Facts
- New Model Support
- GPT-6.2 Astra and Sol
- Headers
- 'langsmith_gateway' headers are now sent
- Client Behavior
- Async HTTP client no longer shared between models

