# LocalAI — Run Any AI Model Locally, No GPU

> LocalAI is an open-source AI engine running LLMs, vision, voice, and image models locally. 44.6K+ GitHub stars. OpenAI/Anthropic-compatible API, 35+ backends, MCP, agents. MIT licensed.

## Install

Save in your project root:

## Quick Use

```bash
# Run with Docker
docker run -p 8080:8080 localai/localai

# Or native install (macOS)
brew install localai

# Run a model from the gallery
local-ai run llama-3.2-1b-instruct:q4_k_m

# Query via OpenAI-compatible API
curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
  "model": "llama-3.2-1b-instruct",
  "messages": [{"role": "user", "content": "Hello!"}]
}'
```

---

## Intro

LocalAI is an open-source AI engine that lets you run LLMs, vision, voice, image, and video models locally without requiring a GPU. With 44,600+ GitHub stars and MIT license, it provides OpenAI and Anthropic-compatible APIs, supports 35+ backends across NVIDIA, AMD, Intel, Apple Silicon, Vulkan, and CPU-only hardware. LocalAI includes enterprise features like authentication, user quotas, RBAC, autonomous AI agents with tool use, RAG, and MCP integration. Run models from HuggingFace, Ollama registries, and OCI images.

**Best for**: Teams wanting a self-hosted AI platform with multi-model support and enterprise features
**Works with**: Claude Code, OpenAI Codex, Cursor, Gemini CLI, Windsurf
**Hardware**: NVIDIA, AMD, Intel, Apple Silicon, Vulkan, CPU-only

---

## Key Features

- **Multi-model**: LLMs, vision, voice, image, video in one platform
- **No GPU required**: Runs on CPU, optional GPU acceleration
- **OpenAI + Anthropic API**: Drop-in replacement for both providers
- **35+ backends**: llama.cpp, Transformers, diffusers, whisper, and more
- **AI agents**: Autonomous agents with tool use, RAG, MCP
- **Enterprise**: Auth, quotas, RBAC, audit logging
- **Model gallery**: One-command install from HuggingFace, Ollama, OCI

---

### FAQ

**Q: What is LocalAI?**
A: LocalAI is a local AI engine with 44.6K+ stars running LLMs, vision, voice, and image models. OpenAI/Anthropic-compatible API, 35+ backends, no GPU required. MIT licensed.

**Q: How do I install LocalAI?**
A: `docker run -p 8080:8080 localai/localai` or `brew install localai` on macOS. Then `local-ai run <model-name>`.

---

## Source & Thanks

> Created by [Ettore Di Giacinto](https://github.com/mudler). Licensed under MIT.
> [mudler/LocalAI](https://github.com/mudler/LocalAI) — 44,600+ GitHub stars

---
Source: https://tokrepo.com/en/workflows/34c0d47e-fb4c-442b-819c-9b6a5f921e13
Author: AI Open Source