Fixing the "--effort is not supported" Error in Gemini 3.1 Pro API Calls

If you are encountering the following error when trying to run your AI models:

Error: error: invalid model selection (--model "gemini-3.1-pro" --effort "low"): --effort is not supported for model "gemini-3.1-pro"

💡 For immediate deployment: The complete source code suite (ZIP) for this architecture is available on Gumroad for $0+ (Pay What You Want).

Here is the quick fix and explanation for why this happens.

Cause of the Error

The --effort flag (often used to control reasoning effort or compute time, similar to OpenAI's reasoning effort parameters) is not supported by the gemini-3.1-pro model endpoint in your current CLI or API configuration.

Solution

Remove the --effort "low" (or any other effort level like --effort "high") parameter from your command or configuration file.

Incorrect Command Example:

ai-tool --model "gemini-3.1-pro" --effort "low" -p "Your prompt here"

Corrected Command Example:

ai-tool --model "gemini-3.1-pro" -p "Your prompt here"

Once you remove the unsupported argument, the model call should execute successfully.

---

If this engineering log saved your production server (and your sanity), consider supporting our architecture on GitHub Sponsors.