Remembers your habits, builds Agent teams, automates the repetitive work.
🏷️ SuperInstance Fork — This is the SuperInstance fork of MiniMax Code, reconfigured with z.ai GLM-5.1 as the primary code generation backend and DeepInfra as fallback. See SUPERINSTANCE_CHANGES.md for details.
This fork replaces MiniMax's default providers with our own stack:
| Role | Provider | Model | Best For |
|---|---|---|---|
| Primary | z.ai | GLM-5.1 | All code tasks |
| Fallback | DeepInfra | seed-2.0-mini | Quick completions |
| Fallback | DeepInfra | gemma-4 | General generation |
| Fallback | DeepInfra | nemotron-120b | Complex algorithms |
| Fallback | DeepInfra | qwen-3.6 | Multilingual code |
| Fallback | DeepInfra | hermes-405b | Large refactoring |
# Set your API keys
export SIA_ZAI_API_KEY=your-zai-key
export SIA_DEEPINFRA_API_KEY=your-deepinfra-key
# Provider config is in config/superinstance-providers.json
# Load it in your MiniMax Code configurationSee SUPERINSTANCE_CHANGES.md for the full changelog and config/superinstance-providers.json for the provider configuration.
This repository collects issue reports for the SuperInstance fork of MiniMax Code.
For bug reports, please include: version, platform, steps to reproduce, and expected vs actual behavior.
- Tap Issues → New issue at the top right
- Pick a template (Bug / Feature / Question)
- Fill in the requested details:
- Version — found in Settings
- Platform — macOS / Windows
- Steps to reproduce — short and reproducible
- Expected behavior — what should happen
- Actual behavior — what actually happens
- Search to see if the issue already exists
- Make sure you are on the latest version
- Strip out any sensitive information (API keys, tokens, etc.)
© 2026 MiniMax. SuperInstance modifications © 2026 SuperInstance. All rights reserved.