[Bug][Copilot Agents] session.create fails with "Model 'gemini-3.8-flash' does not support reasoning effort configuration"

[outwebbed]

🏷️ Discussion Type

Bug

💬 Feature/Topic Area

Copilot Cloud Agent

Summary

When creating or starting a new GitHub Copilot Agent session in a repository using the Gemini 3.8 Flash model, the task terminates immediately at startup during session.create before executing any codebase operations.

---

Steps to Reproduce

1. Navigate to a repository’s Agents tab on GitHub (or open the Copilot Agents workspace).
2. Start a new agent task / session.
3. Select Gemini 3.8 Flash as the model.
4. Submit any prompt (e.g., repository investigation or code review request).
5. Notice the session fails immediately during initialization.

---

Expected Behavior

Copilot Agents should either:

• Detect that gemini-3.8-flash does not support reasoning effort/budget parameters and omit reasoning_effort from the session.create payload, or
• Prevent models lacking reasoning configuration from being selectable in the Agents UI if the agent runtime strictly requires reasoning parameters.

---

Actual Behavior

The session fails immediately upon submission with the following error:

Copilot stopped work due to an error
Request session.create failed with message: Model 'gemini-3.8-flash' does not support reasoning effort configuration. Use models.list to check which models support reasoning effort.

[github-actions[bot]]

💬 Your Product Feedback Has Been Submitted 🎉

Thank you for taking the time to share your insights with us! Your feedback is invaluable as we build a better GitHub experience for all our users.

Here's what you can expect moving forward ⏩

• Your input will be carefully reviewed and cataloged by members of our product teams. 
  • Due to the high volume of submissions, we may not always be able to provide individual responses.
  • Rest assured, your feedback will help chart our course for product improvements.
• Other users may engage with your post, sharing their own perspectives or experiences. 
• GitHub staff may reach out for further clarification or insight. 
  • We may 'Answer' your discussion if there is a current solution, workaround, or roadmap/changelog post related to the feedback.

Where to look to see what's shipping 👀

• Read the Changelog for real-time updates on the latest GitHub features, enhancements, and calls for feedback.
• Explore our Product Roadmap, which details upcoming major releases and initiatives.

What you can do in the meantime 💻

• Upvote and comment on other user feedback Discussions that resonate with you.
• Add more information at any point! Useful details include: use cases, relevant labels, desired outcomes, and any accompanying screenshots.

As a member of the GitHub community, your participation is essential. While we can't promise that every suggestion will be implemented, we want to emphasize that your feedback is instrumental in guiding our decisions and priorities.

Thank you once again for your contribution to making GitHub even better! We're grateful for your ongoing support and collaboration in shaping the future of our platform. ⭐

[alexisvaleev]

This looks like a model configuration compatibility issue.

The error indicates that gemini-3.8-flash does not support reasoning effort configuration, but the Copilot Agent runtime appears to pass reasoning-related parameters during session.create regardless.

Ideally, the runtime should check the selected model's capabilities before constructing the session payload and omit unsupported parameters such as reasoning_effort. Alternatively, models that cannot satisfy the runtime's required configuration should be filtered out of the model selection UI.

It would also be helpful to include the model ID and the relevant session configuration in the error logs to make this easier to diagnose.

Has anyone confirmed whether this is specific to gemini-3.8-flash or affects other models that don't support reasoning effort configuration as well?

[bomtobe-hgkim]

A regression test can pin this boundary: for a model whose capability entry lacks reasoning-effort support, assert that session.create omits the field; for a supported model, assert that it passes the selected value. Run the same contract across model IDs so capability-table drift fails before a task starts. If model switching is user-controlled, surface unsupported options before launch and keep diagnostics to the model ID and config field, without prompts or credentials. Disclosure: I work at BOMTOBE, maker of Orke.