[ENHANCEMENT] add model minimax -m2.5

[ataniii]

Problem (one or two sentences)

minimax m2.5 just came out and using their api there is no selection inside of roo code for it
https://www.minimax.io/news/minimax-m25
model card ^

also maybe consider adding a feature so i can select my own model by text for the api so i dont have to wait for a roo code update

Context (who is affected and when)

anyone using new minimax models

Desired behavior (conceptual, not technical)

make it work

Constraints / preferences (optional)

No response

Request checklist

• I've searched existing Issues and Discussions for duplicates
• This describes a specific problem with clear context and impact

Roo Code Task Links (optional)

No response

Acceptance criteria (optional)

No response

Proposed approach (optional)

No response

Trade-offs / risks (optional)

No response

[roomote]

I have analyzed this issue and here is my proposed implementation plan for adding MiniMax M2.5 support.

What needs to change:

1. packages/types/src/providers/minimax.ts -- Add a "MiniMax-M2.5" entry to minimaxModels following the same pattern as the existing M2, M2-Stable, and M2.1 entries
2. Optionally update minimaxDefaultModelId to "MiniMax-M2.5" if this should become the new default (or keep it as "MiniMax-M2")
3. Update existing tests in src/api/providers/__tests__/minimax.spec.ts if the default model ID changes

No changes are needed to the MiniMax handler itself (src/api/providers/minimax.ts), the provider settings schema, or the UI -- they all dynamically read from the minimaxModels object.

Clarifying questions about model specs:

Since I cannot access the model card URL directly, I need confirmation on these M2.5 parameters:

• Model ID string: Is the API model ID MiniMax-M2.5 (matching the naming convention of MiniMax-M2.1)?
• Max output tokens: Is it 16,384 like the other MiniMax models?
• Context window: Is it 192,000 tokens like M2/M2.1, or has it changed?
• Supports images: Does M2.5 support image inputs?
• Pricing: What are the input/output/cache-write/cache-read prices per million tokens? (M2/M2.1 are $0.30/$1.20/$0.375/$0.03)
• Should M2.5 become the new default model for the MiniMax provider?

If you can confirm these specs (or point me to the pricing page), I will proceed with the implementation. Please mention @roomote in a comment saying that you approve this plan, and I will proceed with the implementation.

[ataniii]

@roomote

here are some specs i found for calls (py and curl)
import requests
url = "https://api.minimax.io/v1/text/chatcompletion_v2"
payload = {
"model": "MiniMax-M2.5",
"messages": [
{"role": "user", "content": "Hello"}
]
}
headers = {"Authorization": "Bearer "}
response = requests.post(url, json=payload, headers=headers)
print(response.text)

curl

curl -X POST "https://api.minimax.io/v1/text/chatcompletion_v2"
-H "Content-Type: application/json"
-H "Authorization: Bearer "
-d '{
"model": "MiniMax-M2.5",
"messages": [
{"role": "user", "content": "Hello"}
]
}'

Should M2.5 become the new default model for the MiniMax provider? i think it should as its in the cheapest code subscription

some links
https://www.minimax.io/models/text
https://www.minimax.io/news/minimax-m25
https://huggingface.co/blog/MiniMax-AI/forge-scalable-agent-rl-framework-and-algorithm

context window 204,800
131,072 maximum output capacity
73,728 tokens for input context
prompt caching at $0.03/M read and $0.375/M write
input 0.30/m output 1.20/m at 204k context

[roomote]

Thanks for the detailed specs @ataniii! I've got everything I need to proceed -- model ID MiniMax-M2.5, 204,800 context window, 131,072 max output, and the pricing details. I'll add it as the new default for the MiniMax provider and open a draft PR shortly.

[roomote]

Opened a draft PR with my attempt at adding MiniMax M2.5: #11458

The PR adds M2.5 as the new default model with the specs from this thread (204K context, 131K max output, same pricing as M2). All existing tests pass. Feedback welcome!