[BUG] The MiniMax-M2.7 model is missing

[antoniodiviggiano]

Problem (one or two sentences)

The MiniMax-M2.7 model is missing

Context (who is affected and when)

The model is missing from the dropdown menu when selecting MiniMax as the provider in the settings.

Reproduction steps

Settings -> Providers -> Model

Expected result

The MiniMax-M2.7 model should be available and selectable in the model list.

Actual result

The model is missing from the list, making it impossible to select or use it within the extension.

Variations tried (optional)

No response

App Version

3.51.1

API Provider (optional)

None

Model Used (optional)

No response

Roo Code Task Links (optional)

No response

Relevant logs or errors (optional)

I've analyzed this issue and here's my proposed implementation plan:

The MiniMax-M2.7 model needs to be added to the model registry. The model definitions live in packages/types/src/providers/minimax.ts, which currently lists: MiniMax-M2.5, MiniMax-M2, MiniMax-M2-Stable, and MiniMax-M2.1.

Proposed changes:

1. Add a "MiniMax-M2.7" entry to the minimaxModels object in packages/types/src/providers/minimax.ts with appropriate configuration (context window, pricing, capabilities, etc. -- matching the pattern of other M2-series models, adjusted for M2.7 specs)
2. Add corresponding test coverage in src/api/providers/__tests__/minimax.spec.ts to verify the new model is correctly configured and selectable

Before I proceed, could you confirm the model specs for MiniMax-M2.7? Specifically:

• Context window size (e.g., 192K like M2, or 204.8K like M2.5, or different?)
• Max output tokens (16,384 like the other models?)
• Input/output pricing (same as other M2-series: $0.30/$1.20 per million tokens?)
• Does it support prompt caching?
• Does it support images?
• Any other notable capabilities?

If the specs match the existing M2-series models (same pricing, 192K context, 16K max tokens, prompt caching, no image support), I can go ahead with those defaults.

Please mention @roomote in a comment saying that you approve this plan, and I'll proceed with the implementation.

[antoniodiviggiano]

I confirm the specs for MiniMax‑M2.7 as follows:

Context window: 204.8K tokens

Max output: 131072 tokens

Pricing: same as the M2‑series: $0.30 per million input tokens and $1.20 per million output tokens

Prompt caching is supported

Images: not supported
You can proceed with the implementation following the M2‑series pattern; I approve this plan.
@roomote please go ahead.

I have seen the confirmed specs and approval. I will proceed with adding MiniMax-M2.7 to the model registry following the M2-series pattern. Will open a PR shortly.

Opened a draft PR with MiniMax-M2.7 added to the provider registry: #11982. Feedback welcome!

[luguoyixiazi]

Opened a draft PR with MiniMax-M2.7 added to the provider registry: #11982. Feedback welcome!

Shouldn't use hard coding，use api to get model info or let user input info，this is the most important PR that's need to be merged，merged now

[RysanekDavid]

Can I ask when the release is scheduled so that Minmax 2.7 will finally be available? Thanks

[tea0112]

I am really in the need of M2.7

[childishsue]

Has this problem been fixed? I am really in the need of M2.7 too

[Leewaldoe]

Interested in this as well.

[vitalimi]

Please add model M2.7