Skip to content

Default minimax provider to MiniMax-M3 in setup wizard#83

Open
octo-patch wants to merge 1 commit into
aiming-lab:mainfrom
octo-patch:feature/upgrade-minimax-m3
Open

Default minimax provider to MiniMax-M3 in setup wizard#83
octo-patch wants to merge 1 commit into
aiming-lab:mainfrom
octo-patch:feature/upgrade-minimax-m3

Conversation

@octo-patch
Copy link
Copy Markdown
Contributor

Summary

The minimax preset in metaclaw/setup_wizard.py shipped with an empty model_id default, forcing first-time users to look up a valid model ID before the wizard could finish. The other API-key presets that have a clear flagship (kimi, qwen, openai, volcengine) all ship one, so this brings minimax in line.

MiniMax-M3 is the current flagship MiniMax model and is the right default for a fresh metaclaw setup. Users who want a different model can still override it at the wizard's Model ID prompt or via metaclaw config llm.model_id <id> afterwards.

Changes

  • metaclaw/setup_wizard.py: set the default model_id for the minimax preset from "" to "MiniMax-M3". No changes to api_base, no changes to any other provider.

Notes

  • One-line, opt-out change — the wizard still prompts the user, so explicit overrides keep working.
  • No changes to API base URL or to other providers.
  • No new dependencies, no test changes (the existing test_setup_wizard.py mocks the model ID prompt so it is unaffected).

The minimax preset previously left model_id empty, forcing every user
to look up a model name during the first-time setup. Set the default
to the current flagship model `MiniMax-M3` so the wizard runs end to
end without manual lookup, matching how the kimi/qwen/openai/volcengine
presets already ship a sensible default. Users can still override it
at the wizard's "Model ID" prompt or via `metaclaw config llm.model_id`.

Co-Authored-By: Octopus <liyuan851277048@icloud.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant