Skip to content

Support configurable reasoning effort for Responses API Bots #212

Description

@arcadegeek23

Description

Allow the LangGraph/Responses API Bot to configure reasoning effort through deployment configuration and use it with GPT-5.6 model IDs such as gpt-5.6-luna.

Requested behavior

  • Add an environment setting for reasoning effort.
  • Validate supported values rather than silently ignoring invalid configuration.
  • Pass the setting to the OpenAI Responses API through the LangChain OpenAI integration.
  • Document the setting and add regression coverage.
  • Keep the hand-written chat-completions Bot on a compatible model; GPT-5.6 must use the Responses API Bot.

Security

No credentials are included in this issue. API keys remain deployment-local and must not be committed.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions