LocalAI

A self-hosted, open-source alternative to OpenAI's API. Run large language models, generate images, audio transcription, and embeddings on your own infrastructure. Compatible with the OpenAI API specification for seamless integration with existing tools and libraries.

What You Can Do After Deployment

1. Visit your domain — verify the LocalAI web UI is running
2. Download models — browse and install models from the built-in gallery
3. Use the OpenAI-compatible API — point any OpenAI SDK client at your LocalAI instance
4. Generate text — run chat completions and text generation with local models
5. Generate images — use Stable Diffusion models for image generation
6. Transcribe audio — use Whisper models for speech-to-text

Key Features

• OpenAI API compatible endpoints (chat, completions, embeddings, images, audio)
• Built-in model gallery with one-click downloads
• Runs on CPU — no GPU required (GPU acceleration optional)
• Supports multiple model formats (GGUF, GGML, transformers)
• Text-to-speech and speech-to-text capabilities
• Image generation with Stable Diffusion backends
• Function calling and tool support
• REST API and gRPC backends
• Web UI for model management and chat

License

MIT — GitHub | Website