MiniMax (coding) for VS Code

Language model chat provider for GitHub Copilot in VS Code using MiniMax text models with a Token Plan API key.

Features

• Token Plan API key from platform.minimax.io
• OpenAI-compatible chat to https://api.minimax.io/v1
• Tool calling and reasoning/thinking streaming
• M3 model supports image input (multimodal)

Requirements

• VS Code 1.111.0+
• MiniMax Token Plan subscription and API key
• VS Code Insiders is required to render MiniMax thinking blocks via the proposed languageModelThinkingPart API

Setup

1. Get your Token Plan API key from Account / Token Plan
2. Use the API key navigation action in the model picker
3. Choose a model in the Copilot model picker

Keys are stored in VS Code Secret Storage.

Configuration

minimax.visibleModels (array of model IDs) controls which models appear in the picker.

Models

Model	Context	Max input	Max output
MiniMax-M3	1,000,000	1,000,000	131,072
MiniMax-M2.7	204,800	200,000	131,072
MiniMax-M2.7-highspeed	204,800	200,000	131,072
MiniMax-M2.5	204,800	196,000	128,000
MiniMax-M2.5-highspeed	204,800	196,000	128,000
MiniMax-M2.1	204,800	196,000	128,000
MiniMax-M2.1-highspeed	204,800	196,000	128,000
MiniMax-M2	204,800	192,000	128,000

License

MIT