Skip to main content

EASY AS PIE UPDATE: Support reasoning effort = NONE, low, medium, high for Gemini in accordance to new feature in Gemini Openai compatibility layer released Apr 27, 2025

Feature Request

https://x.com/OfficialLoganK/status/1915934863527420398/photo/1

Unlike the Gemini API, the OpenAI API offers three levels of thinking control: "low", "medium", and "high", which behind the scenes we map to 1K, 8K, and 24K thinking token budgets.

If you want to disable thinking, you can set the reasoning effort to "none".

https://ai.google.dev/gemini-api/docs/openai

Status: Completed6 comments

Log in to comment and vote

Comments6

  • Amaranth Bottle

    •

    May 20, 2025

    Can you make it so that the model settings page for Gemini 2.5 Flash has the "Reasoning Effort" setting as well?

  • Ngoc Nguyen changed status to Completed
    Team•

    May 13, 2025

    this is now available for Gemini 2.5 Flash and Claude 3.7 (for now)

    • Rose Pudding

      •

      May 23, 2025

      Ngoc Nguyen can you guys add this also to the parameter tab for each model instead of only have this global settings?

  • Ngoc Nguyen changed status to In Progress
    Team•

    May 12, 2025

  • Coral Deer

    •

    Apr 28, 2025

    i believe this is only for the Flash model if im correct

  • Rose Pudding

    •

    Apr 24, 2025

    Yes, also 2.5 flash with reasoning enabled is 6x the price vs not enabled