Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Allow configuration of Reasoning_Effort on modelSettings #189

Closed
exiao opened this issue Mar 17, 2025 · 1 comment · Fixed by #388 · May be fixed by #235
Closed

Allow configuration of Reasoning_Effort on modelSettings #189

exiao opened this issue Mar 17, 2025 · 1 comment · Fixed by #388 · May be fixed by #235
Labels
enhancement New feature or request good first issue Good for newcomers

Comments

@exiao
Copy link

exiao commented Mar 17, 2025

Reasoning models are the best for getting the most accurate result, but they take forever. I'd like to specify the reasoning_effort to be low, like such:

agent = Agent(
        name="",
        instructions="",
        model="o3-mini",
        model_settings=ModelSettings(parallel_tool_calls=True, truncation="auto", reasoning_effort="low"),
        tools=[]
)

https://platform.openai.com/docs/guides/reasoning?api-mode=responses

But it is not available on ModelSettings
https://openai.github.io/openai-agents-python/ref/model_settings/

This is model_settings.py
Image

@exiao exiao added the enhancement New feature or request label Mar 17, 2025
@rm-openai rm-openai added the good first issue Good for newcomers label Mar 17, 2025
@farouqalsalih
Copy link

Hello, can I take this as my first github issue?

askmuhsin pushed a commit to askmuhsin/openai-agents-python that referenced this issue Mar 19, 2025
rm-openai added a commit that referenced this issue Apr 3, 2025
fixes #189 

@rm-openai Would really appreciate if this can get a quick review.

---------

Co-authored-by: Rohan Mehta <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request good first issue Good for newcomers
Projects
None yet
3 participants