feat: add MiniMax as a supported LLM provider

[octo-patch]

Summary

• Add MiniMax as a first-class LLM provider for Quivr
• Register MiniMax-M2.5 and MiniMax-M2.5-highspeed models (204K context window, up to 192K output tokens)
• Use MiniMax's OpenAI-compatible API endpoint (https://api.minimax.io/v1)
• Add unit test for MiniMax endpoint creation
• Update README to list MiniMax among supported providers

Details

MiniMax offers OpenAI-compatible chat completion APIs. This PR adds MiniMax as a dedicated supplier in Quivr's DefaultModelSuppliers enum, with proper model configs and endpoint handling via ChatOpenAI.

Supported Models

Model	Context Window	Max Output	Description
MiniMax-M2.5	204,800 tokens	192,000 tokens	Peak Performance. Ultimate Value.
MiniMax-M2.5-highspeed	204,800 tokens	192,000 tokens	Same performance, faster and more agile

Usage

export MINIMAX_API_KEY=your_api_key

from quivr_core.rag.entities.config import DefaultModelSuppliers, LLMEndpointConfig
from quivr_core.llm import LLMEndpoint

config = LLMEndpointConfig(
    supplier=DefaultModelSuppliers.MINIMAX,
    model="MiniMax-M2.5",
)
llm = LLMEndpoint.from_config(config)

Test plan

• Syntax validation passes for all modified files
• MiniMax API connectivity verified with live API call
• Unit test added for MiniMax endpoint creation
• Full test suite (requires Python 3.11+)

Files changed

• core/quivr_core/rag/entities/config.py - Add MINIMAX enum + model configs
• core/quivr_core/llm/llm_endpoint.py - Add MiniMax handler
• core/tests/test_llm_endpoint.py - Add MiniMax test
• README.md - Update supported providers list

[cubic-dev-ai]

No issues found across 3 files

[cubic-dev-ai]

1 issue found across 2 files (changes from recent commits).

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="core/quivr_core/rag/entities/config.py">

<violation number="1" location="core/quivr_core/rag/entities/config.py:280">
P2: `"MiniMax-M2.7-highspeed"` config is unreachable via `get_llm_model_config` because `"MiniMax-M2.7"` appears first in the dict and matches via `startswith`. Swap the order so the longer (more specific) key is checked first.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}

[octo-patch]

Good catch! Swapped the dict ordering so MiniMax-M2.7-highspeed appears before MiniMax-M2.7 (and similarly MiniMax-M2.5-highspeed before MiniMax-M2.5), ensuring the more specific key is matched first via startswith. Thanks for the review!