Skip to content

Releases: derekwin/ModelGate

v0.0.1

28 Feb 07:09
@derekwin derekwin
v0.0.1
6806b56

Choose a tag to compare

View all tags
v0.0.1 Latest
Latest

🚀 ModelGate v0.0.1

We are excited to announce the first official release of ModelGate — an OpenAI-compatible API Gateway for local LLM deployment.

ModelGate is designed to solve the governance, security, and operational challenges of running local large language models in production environments.

✨ Highlights

🔐 Secure API Key Management

  • SHA256-hashed API key storage
  • Per-key IP whitelist support
  • Admin key hot-reload support

📊 Token Quota & Usage Tracking

  • Per-user token quota control
  • Real-time token consumption accounting
  • Full request usage logging

⚡ Rate Limiting

  • Redis-backed rate limiting
  • Configurable RPM and burst control
  • Designed for multi-tenant environments

🌐 Multi-Backend Support

  • Ollama
  • vLLM
  • llama.cpp
  • OpenAI (hybrid cloud support)
  • API3 integration

Adapter-based backend design enables easy extension for new model providers.

🔄 OpenAI API Compatibility

  • Fully compatible with OpenAI API format
  • Zero-cost migration (only base_url change required)
  • Supports /v1/chat/completions and standard endpoints

🎛 Flexible Management Interfaces

  • Web Admin UI
  • CLI management tool
  • RESTful admin APIs

🎯 Use Cases

  • Enterprise internal AI governance
  • SaaS tiered AI service routing
  • Hybrid local + cloud deployment
  • Vertical model API productization
  • Multi-instance model load balancing

🙌 Acknowledgements

Thanks to all early users and contributors who provided feedback during development.

If you find ModelGate useful, please consider ⭐ starring the repository.

Contributions, issues, and feature requests are welcome.

Assets 2
Loading