Skip to content

feat: add MiniMax M2.7 and M2.5-highspeed models#170

Closed
octo-patch wants to merge 1 commit intoTHU-MAIC:mainfrom
octo-patch:feature/add-minimax-m27-model
Closed

feat: add MiniMax M2.7 and M2.5-highspeed models#170
octo-patch wants to merge 1 commit intoTHU-MAIC:mainfrom
octo-patch:feature/add-minimax-m27-model

Conversation

@octo-patch
Copy link

Summary

  • Add MiniMax-M2.7 (latest flagship) and MiniMax-M2.5-highspeed (fast variant) to the built-in MiniMax provider
  • List MiniMax in the supported providers section of both English and Chinese READMEs
  • Add vitest testing framework with 28 unit + integration tests for provider configuration

Changes

File Change
lib/ai/providers.ts Add M2.7 and M2.5-highspeed models to MiniMax provider
README.md Add MiniMax to supported providers list
README-zh.md Add MiniMax to supported providers list (Chinese)
vitest.config.ts New vitest configuration
package.json Add vitest devDependency and test scripts
lib/ai/__tests__/providers.test.ts 21 unit tests for MiniMax provider config
lib/ai/__tests__/providers.integration.test.ts 7 integration tests for getModel()

Context

MiniMax recently released M2.7, their latest flagship model. This PR adds it as the top model in the existing MiniMax provider, along with the M2.5-highspeed variant for latency-sensitive workloads. The Anthropic-compatible API endpoint (api.minimaxi.com/anthropic/v1) supports all MiniMax models.

Test plan

  • All 28 vitest tests pass (npx vitest run)
  • Verify M2.7 model works via MiniMax Anthropic-compatible API
  • Verify model appears in the provider/model selector UI

Add the latest MiniMax-M2.7 flagship model and MiniMax-M2.5-highspeed
to the built-in MiniMax provider. Update both English and Chinese READMEs
to include MiniMax in the supported providers list. Add vitest with unit
and integration tests for the MiniMax provider configuration.
@wyuc
Copy link
Contributor

wyuc commented Mar 24, 2026

Thanks for the contribution. PR #182 is adding broader MiniMax support (LLM + TTS + Image + Video) and we're using that as the main MiniMax PR. I've asked the author there to include your MiniMax-M2.7 model entry as well.

If you'd prefer, you can also open a smaller PR after #182 lands with just the model addition. Either way your work won't be lost.

A couple of notes if you do update this PR:

@octo-patch
Copy link
Author

Thanks for the heads up, @wyuc! PR #182 sounds like the better fit for broader MiniMax coverage (LLM + TTS + Image + Video). I'll close this PR. Happy to open a follow-up after #182 lands if any M2.7 model entries are still needed.

@octo-patch octo-patch closed this Mar 24, 2026
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.

2 participants