Recommended Open-Source AI Projects
A hand-picked list of open-source AI projects you can self-host or run locally — covering local inference, app orchestration, foundation models, and multimodal — to support your private AI and DevOps practice.
Featured Projects
Ollama
FeaturedRun LLMs locally with a single command (Llama, Qwen, and more) — a minimal local LLM runtime.
vLLM
FeaturedHigh-throughput LLM inference engine (PagedAttention) for production-grade serving.
Open WebUI
FeaturedOpen-source AI application UI for local models — one-click hookup to Ollama/vLLM, ChatGPT-like experience.
Dify
FeaturedOpen-source LLM application development platform (Agents, visual workflows, RAG).
RAGFlow
FeaturedOpen-source deep document retrieval (RAG) engine, well suited for private knowledge bases.
LangChain
FeaturedThe foundational framework and component library for building LLM applications.
LlamaIndex
FeaturedA data framework for wiring private data and documents into LLM apps.
Qwen
FeaturedAlibaba's open-source Qwen model family (including reasoning and multimodal).
DeepSeek
FeaturedDeepSeek's open-source LLMs (the DeepSeek-V3 / R1 reasoning series).
Llama
FeaturedMeta's open-source Llama foundation model series.
Whisper
FeaturedOpenAI's open-source speech recognition (ASR) — multilingual and robust.