bug: O1-Preview does not work because of: "Unsupported parameter: 'max_tokens' is not supported with this model.".

[raphaelmosaic]

Jan version

v0.5.4-640

Describe the Bug

When making a request using o1-preview or o1-mini, getting:

Unsupported parameter: 'max_tokens' is not supported with this model. Use 'max_completion_tokens' instead. Jan’s in beta. Access troubleshooting assistance now.

Cannot change these model parameters settings.

Steps to Reproduce

No response

Screenshots / Logs

What is your OS?

• [X] MacOS
• [ ] Windows
• [ ] Linux

[imtuyethan]

Thanks for reporting this @raphaelmosaic.

https://community.openai.com/t/why-was-max-tokens-changed-to-max-completion-tokens/938077/3

Looks like the new o1 series models (like o1-preview and o1-mini) use max_completion_tokens instead of max_tokens. We'll work on updating Jan to handle this smoothly.

For now, here's a quick workaround:

1. Navigate to your Jan data folder (usually in your user directory)
2. Find models folder, then find the models/o1-mini/model.json file (and o1-preview if you use that too)
3. Open it in a text editor
4. Change max_tokens to max_completion_tokens
5. Save the file and restart Jan

This should get you up and running. We'll be pushing an update soon to handle this automatically.

[raphaelmosaic]

Changed this, still does not work. Please fix this properly asap as I am 100% one of many people who would use o1mini / o1preview via Jan and have this issue

[louis-jan]

Build containing bug fixes: https://github.com/janhq/jan/actions/runs/11007851785