Captures d'écran du projet
Vue d’ensemble
AnythingLLM is a comprehensive, self-hostable AI application that allows users to build a private, fully-featured AI assistant. It supports connecting various local or cloud LLMs, ingesting multiple document types, and chatting with data using built-in RAG (Retrieval-Augmented Generation) capabilities. The application is designed to be hyper-configurable and production-ready, featuring built-in agents, multi-user support, vector database management, and document pipelines without requiring complex external configurations. It is actively developed by Mintplex Labs and is available as a desktop application or via Docker for self-hosting.
Fonctionnalités clés
- Dynamic Model Routing to automatically select the best provider and model
- Automatic and user-managed LLM memories
- Scheduled tasks with full agent capabilities on a cron schedule
- Intelligent skill selection to reduce token usage by up to 80% per query
- No-code AI Agent builder
- MCP (Model Context Protocol) compatibility
- Multi-modal support for both closed and open-source LLMs
- Multi-user instance support with permissioning (Docker only)
- Custom embeddable chat widget for websites (Docker only)
- Support for multiple document types (PDF, TXT, DOCX, etc.)
- Full Developer API for custom integrations
Prérequis, installation et démarrage rapide
Utilisation
Compatibilité des modèles et cas d’usage
Supports a wide range of LLMs including OpenAI, Azure OpenAI, Anthropic, AWS Bedrock, Google Gemini Pro, NVIDIA NIM, Ollama, LM Studio, LocalAI, Together AI, Fireworks AI, Perplexity, OpenRouter, DeepSeek, Mistral, Groq, Cohere, KoboldCPP, LiteLLM, Text Generation Web UI, Apipie, xAI, Z.AI, Novita AI, PPIO, Gitee AI, Moonshot AI, Microsoft Foundry Local, CometAPI, Docker Model Runner, PrivateModeAI, SambaNova Cloud, Lemonade by AMD, Minimax, Cerebras, oMLX, and any open-source llama.cpp compatible model. Embedder models supported include AnythingLLM Native Embedder, OpenAI, Azure OpenAI, Gemini, LocalAI, Ollama, LM Studio, Lemonade, OpenRouter, LiteLLM, Cohere, Voyage AI, Mistral, and generic OpenAI-compatible APIs. Vector databases supported include LanceDB (default), PGVector, Astra DB, Pinecone, Chroma, Weaviate, Qdrant, Milvus, and Zilliz.
Licence et notes sur les risques
The repository is licensed under the MIT License.
Editorial verification 2026-08-09: repository URL, owner, description, license and repository statistics were reviewed. License metadata: MIT. README was fetched for the channel draft; re-check repository dependencies, releases and model terms before production use.
Publication et maintenance
Not stated in the repository metadata.