Ross ROSS = Recommend OSS · open-source software intelligence for agents

mudler/LocalAI

LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required. observed · 2026-08-28

github.com/mudler/LocalAI · homepage · Go · MIT (permissive) observed · 2026-08-28

Health v2 · maintenance only

93/100

  • Activity 99
  • Release rhythm 86
  • Longevity 90
How is this computed?

round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-03. Adoption (stars, forks) is never an input.

  • gap_med: 2.0
  • age_days: 1264
  • days_rel: 13
  • days_push: 7
  • n_releases_24m: 77

Full methodology

Adoption not part of the score

48696 stars · 4393 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

LocalAI is an open-source, self-hosted AI inference runtime that runs LLMs, vision, speech, image, and video models behind OpenAI/Anthropic/ElevenLabs-compatible APIs on any hardware, including CPU-only machines. It uses a small core with swappable backends (llama.cpp, vLLM, whisper.cpp, stable-diffusion, MLX) pulled on demand, plus built-in agents, RAG, and MCP support.

Use cases

  • run llm locally without gpu
  • self-hosted openai api replacement
  • serve whisper transcription on my own server
  • generate images with stable diffusion locally
  • local text to speech with voice cloning
  • run ai agents with tools and rag on-premise
  • drop-in openai compatible endpoint for existing apps
  • private ai inference that keeps data on my infrastructure

When to choose

  • you need OpenAI-compatible APIs but want models running on your own hardware
  • you have no GPU or mixed hardware (CPU, Apple Silicon, AMD, Intel, NVIDIA)
  • you want one runtime serving text, vision, speech, and image models
  • privacy requires that data never leaves your infrastructure
  • you want built-in agents, RAG, and MCP without extra services

When to avoid

  • you need maximum single-model throughput from a dedicated vLLM/TensorRT deployment
  • you only use one cloud-hosted model and don't need self-hosting
  • you need managed scaling and zero ops responsibility
  • your stack depends on provider-specific features not covered by the compatible APIs

Facets

service · maturity active

llm-inference http-server api-framework tts speech-recognition image-processing computer-vision rag agent-framework mcp chatbot sdk artificial-intelligence large-language-models machine-learning self-hosted speech-processing computer-vision self-hosted windows go cli openai-compatible-api local-inference cpu-inference stable-diffusion whisper llama-cpp vllm voice-cloning realtime-api privacy-first model-gallery drop-in-replacement ai-agents retrieval-augmented-generation docker linux macos web-server gpu

10 sources

Member repositories

RepositoryRoleHealth v2
mudler/LocalAImain93

For agents

markdown · JSON · MCP: product_card(name="mudler/LocalAI")

Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem