All tasks / Tools
LocalAI
Open-source, self-hosted OpenAI-compatible API server that runs LLMs and other models via llama.cpp, vLLM, MLX and more backends.
- Ettore Di Giacinto (mudler)
- Local only
- Free, open source
Where it ranks
- SChat & assistantsRuntimes & gateways · Local · via Kimi K3
- BCodingRuntimes & gateways · Local · via DeepSeek-V4.1-Flash-1017
- AImage understandingRuntimes & gateways · Local · via Gemma 4 (31B)
Tools without a benchmark of their own inherit the tier of the best model they offer.
Facts
- Vendor
- Ettore Di Giacinto (mudler)
- Approach
- Runtimes & gateways
- Where it runs
- Runs only on your own hardware
- Interface
- API, Self-hosted web UI, Command line
- Skill level
- Advanced: command line, node graphs or self-hosting
- Price
- Free, open source
- Pricing notes
- Free and open source.
- Source code
- Open source (MIT)
- GitHub stars
- 49.6k (19 Oct 2026)
- Last verified
- 19 Oct 2026
Price from github.com, 19 Oct 2026.
Models it gives you
- SKimi K3Moonshot AI · open weights
- AGLM-5.3Z.ai · open weights
- ADeepSeek-V4.1-Flash-1017DeepSeek · open weights
- ADeepSeek-V4-Pro-0813DeepSeek · open weights
- AGemma 4 (31B)Google DeepMind · open weights
- AMiniMax-M3MiniMax · open weights
- AQwen3.8-27BAlibaba (Qwen) · open weights
- Bgpt-oss-120bOpenAI · open weights
- BQwen3.8-2.4T-A95BAlibaba (Qwen) · open weights
- CLlama 4 MaverickMeta · open weights
- Cgpt-oss-20bOpenAI · open weights
- CQwen3-Coder-NextAlibaba (Qwen) · open weights
- CDevstral 2 (2512)Mistral AI · open weights
- –Devstral Small 2 (2512)Mistral AI · open weights