Skip to content

feat(provider): add MiniMax M3 model and prune deprecated M2.x variants#3435

Closed
octo-patch wants to merge 1 commit into
tailcallhq:mainfrom
octo-patch:feature/upgrade-minimax-m3
Closed

feat(provider): add MiniMax M3 model and prune deprecated M2.x variants#3435
octo-patch wants to merge 1 commit into
tailcallhq:mainfrom
octo-patch:feature/upgrade-minimax-m3

Conversation

@octo-patch
Copy link
Copy Markdown

Summary

Updates the MiniMax provider model catalog in crates/forge_repo/src/provider/provider.json:

  • Add MiniMax-M3 as the new top-of-list (default) entry — the latest generation MiniMax model with enhanced reasoning, function calling, and an extended (1M) context window.
  • Keep MiniMax-M2.7 and MiniMax-M2.7-highspeed as legacy options for users still on the previous generation.
  • Remove the deprecated entries: MiniMax-M2, MiniMax-M2.1, MiniMax-M2.1-highspeed, MiniMax-M2.5, MiniMax-M2.5-highspeed. These are no longer recommended for new sessions; the M2.7 line covers the same use cases at parity or better quality.

This follows the same minimal-edit pattern used for prior model catalog updates (e.g. #2727, #2701, #2597, #2557). Other knobs — base URL (api.minimax.io), MINIMAX_API_KEY env var, the OpenRouter SetMinimaxParams transformer, and OpenCode model routing — are intentionally untouched.

Why

MiniMax-M3 is the current production model at MiniMax and outperforms the M2.x line on coding/agentic benchmarks. Surfacing it as the default lets new users pick it up automatically; pruning the older entries keeps the picker focused on the models that are still recommended.

Changes

  • crates/forge_repo/src/provider/provider.json — only the id: "minimax" models array is touched. 9 insertions / 49 deletions, all data-only (no schema changes).

Test plan

  • cargo check -p forge_repo passes (data-only edit; existing ProviderConfig schema is unchanged).
  • cargo test -p forge_repotest_load_provider_configs still loads the catalog; no minimax-specific assertions exist that depend on the removed model IDs.
  • Manual: forge lists MiniMax-M3 first under the MiniMax provider; selecting it routes through the existing Anthropic-compatible endpoint without code changes.

Add MiniMax-M3 as the new default model and remove the deprecated M2,
M2.1, M2.1-highspeed, M2.5, and M2.5-highspeed entries. Keeps M2.7 and
M2.7-highspeed as legacy options for users still on the previous
generation.
@CLAassistant
Copy link
Copy Markdown

CLA assistant check
Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.


octo-patch seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
You have signed the CLA already but the status is still pending? Let us recheck it.

@github-actions github-actions Bot added type: feature Brand new functionality, features, pages, workflows, endpoints, etc. type: provider Updates provider.json configuration. labels Jun 2, 2026
@amitksingh1490
Copy link
Copy Markdown
Contributor

closing in favour of #3434

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

type: feature Brand new functionality, features, pages, workflows, endpoints, etc. type: provider Updates provider.json configuration.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants