Adds the newly released MiniMax-M3 flagship model to the minimax provider.

MiniMax just released M3 — the latest M-series model for agentic reasoning, tool use, coding, and long-context tasks.

What changed

Added MiniMax-M3 as the first model in the minimax provider:

{
  "id": "MiniMax-M3",
  "name": "MiniMax-M3",
  "version": "1.0",
  "description": "Latest M-series flagship for agentic reasoning, tool use, coding, and long-context tasks. 1M context window with image and video input.",
  "capabilities": ["completion", "tools", "vision", "reasoning"]
}

Details

• Context window: 1,000,000 tokens (vs. 204K on M2.x).
• Capabilities: unlike the M2.x series (completion + tools only), M3 supports image/video input and reasoning blocks, so vision and reasoning are added.
• Source: https://platform.minimax.io/docs/api-reference/text-anthropic-api

Validation

• ✅ providers/registry.json is valid JSON.
• ✅ Passes schema validation via ajv against providers/schema.json.
• ✅ updated_at bumped.

🤖 Generated with Claude Code