Open-source AI chat apps and local LLM tools

Open-source apps for chatting with AI models, running language models on your own hardware and building private assistants over your documents.

36 open-source projects, checked against GitHub on 1 November 2026.

All 36 AI chat & local LLMs projects

Showing all 36, ordered by health score.

  1. 94 Active
    AnythingLLM

    Stop renting your intelligence. Own it with AnythingLLM. Everything you need for a powerful local-first agent experience

    MIT
    Setup
    1 of 5
    Stars
    67.8k
    Release
    26 October 2026
  2. 93 Active
    RAGFlow

    RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for…

    Apache-2.0 paid features in repo
    Setup
    1 of 5
    Stars
    93.3k
    Release
    30 October 2026
  3. 93 Active
    KoboldCpp

    KoboldCpp - Run GGUF AI models locally with a KoboldAI UI. One File. Zero Install.

    AGPL-3.0
    Setup
    1 of 5
    Stars
    12.1k
    Release
    31 October 2026
  4. 89 Active
    Ollama

    Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

    MIT
    Setup
    Docker
    Stars
    186k
    Release
    28 October 2026
  5. 84 Active
    llama.cpp

    LLM inference in C/C++

    MIT
    Setup
    Builds
    Stars
    133k
    Release
    28 October 2026
  6. 84 Active
    vLLM

    A high-throughput and memory-efficient inference and serving engine for LLMs

    Apache-2.0
    Setup
    Docker
    Stars
    96k
    Release
    18 October 2026
  7. 84 Active
    DocsGPT

    Private AI platform for agents, assistants and enterprise search. Built-in Agent Builder, Deep research, Document analysis, Multi-model support, and API connectivity…

    MIT
    Setup
    3 of 5
    Stars
    18.4k
    Release
    25 October 2026
  8. 83 Active
    LiteLLM

    The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging…

    MIT paid features in repoCLA
    Setup
    2 of 5
    Stars
    62.5k
    Release
    28 October 2026
  9. 82 Active
    FastGPT

    FastGPT is a knowledge-based platform built on the LLMs, offers a comprehensive suite of out-of-the-box capabilities such as data processing, RAG retrieval, and visual…

    FastGPT Open Source License source-available
    Setup
    4 of 5
    Stars
    30k
    Release
    9 October 2026
  10. 82 Active
    MaxKB

    MaxKB is an open-source platform for building enterprise-grade agents. 强大易用的开源企业级智能体平台。

    GPL-3.0
    Setup
    Docker
    Stars
    23.1k
    Release
    22 October 2026
  11. 82 Active
    Local Deep Research

    ~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private…

    MIT
    Setup
    2 of 5
    Stars
    9.5k
    Release
    28 August 2026
  12. 81 Active
    Cherry Studio

    Your AI workspace for models, agents, and productivity.

    AGPL-3.0
    Setup
    Docker
    Stars
    53.2k
    Release
    31 October 2026
  13. 80 Active
    Harbor

    Stop configuring your AI stack. Start using it. One command brings a complete pre-wired LLM stack with hundreds of services to explore.

    Apache-2.0
    Setup
    1 of 5
    Stars
    3.3k
    Release
    31 October 2026
  14. 79 Active
    OpenClaw

    The AI that really does things. Any OS. Any Platform. The lobster way.

    MIT
    Setup
    3 of 5
    Stars
    394k
    Release
    29 October 2026
  15. 79 Active
    Chatbox

    Powerful AI Client

    GPL-3.0 relicensed 2024
    Setup
    Docker
    Stars
    42.2k
    Release
    24 September 2026
  16. 77 Active
    Onyx

    Open Source AI Platform - AI Chat with advanced features that works with every LLM

    MIT paid features in repo
    Setup
    5 of 5
    Stars
    33.1k
    Release
    30 October 2026
  17. 76 Active
    Hermes Agent

    The agent that grows with you

    MIT
    Setup
    2 of 5
    Stars
    259k
    Release
    23 October 2026
  18. 76 Active
    Open WebUI

    User-friendly AI Interface (Supports Ollama, OpenAI API, ...)

    Open WebUI License source-availablerelicensed 2025CLA
    Setup
    2 of 5
    Stars
    157k
    Release
    29 October 2026
  19. 76 Active
    Open Notebook

    An Open Source implementation of Notebook LM with more flexibility and features

    MIT
    Setup
    2 of 5
    Stars
    41.2k
    Release
    4 October 2026
  20. 75 Active
    LocalAI

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

    MIT
    Setup
    1 of 5
    Stars
    50.1k
    Release
    23 October 2026
  21. 74 Active
    LibreChat

    Enhanced ChatGPT Clone: Features Agents, MCP, Skills, DeepSeek, Anthropic, AWS, OpenAI, Responses API, Azure, Groq, o1, GPT-5, Mistral, OpenRouter, Vertex AI, Gemini,…

    MIT
    Setup
    3 of 5
    Stars
    46.6k
    Release
    26 October 2026
  22. 74 Active
    GPT Researcher

    An autonomous agent that conducts deep research on any data using any LLM providers

    Apache-2.0 relicensed 2024
    Setup
    3 of 5
    Stars
    30.4k
    Release
    26 September 2026
  23. 74 Active
    Page Assist

    Use your locally running AI models to assist you in your web browsing

    MIT
    Setup
    From source
    Stars
    8.3k
    Release
    25 October 2026
  24. 73 Active
    LobeHub

    LobeHub is your Chief Agent Operator, organizing your agents into 7×24 operations by hiring, scheduling, and reporting on your entire AI team.

    LobeHub Community License source-availablerelicensed 2024
    Setup
    5 of 5
    Stars
    83.7k
    Release
    1 November 2026
  25. 73 Active
    Jan

    Jan is an open source alternative to ChatGPT that runs 100% offline on your computer.

    Apache-2.0 relicensed 2025
    Setup
    Builds
    Stars
    45.2k
    Release
    28 October 2026
  26. 71 Active
    big-AGI

    AI suite powered by state-of-the-art models and providing advanced AI/AGI functions. Includes AI personas, AGI functions, world-class Beam multi-model chats,…

    MIT
    Setup
    1 of 5
    Stars
    7.2k
    Release
    27 October 2026
  27. 67 Active
    PrivateGPT

    Complete API layer for private AI applications on local models: RAG, skills, tools, MCP, text-to-sql, and more. Works with any OpenAI-compatible inference server.

    Apache-2.0
    Setup
    Docker
    Stars
    57.6k
    Release
    18 June 2026
  28. 62 Active
    SillyTavern

    LLM Frontend for Power Users.

    AGPL-3.0
    Setup
    1 of 5
    Stars
    35.2k
    Release
    14 September 2026
  29. 58 Active
    llamafile

    Distribute and run LLMs with a single file.

    Apache-2.0
    Setup
    From source
    Stars
    26.5k
    Release
    15 September 2026
  30. 57 Active
    Leon

    Leon is your open-source personal assistant.

    MIT
    Setup
    Builds
    Stars
    17.6k
    Release
    19 February 2026
  31. 43 Active
    exo

    Run frontier AI locally.

    Apache-2.0 relicensed 2025
    Setup
    Builds
    Stars
    48.1k
    Release
    23 April 2026
  32. 27 Active
    NextChat

    Zero-config AI chat assistant. No API key needed — sign up and instantly chat with GPT-5, Claude 4, Gemini 2.5, DeepSeek & 100+ top models. Pay-as-you-go saves you…

    MIT
    Setup
    2 of 5
    Stars
    88.9k
    Release
    29 July 2025
  33. 26 Active
    Text Generation WebUI

    Open-source desktop app for local LLMs. Text, vision, tool-calling, OpenAI/Anthropic-compatible API. 100% private.

    AGPL-3.0
    Setup
    1 of 5
    Stars
    47.7k
    Release
    20 May 2026
  34. 19 Slowing
    Vane

    Vane is an AI-powered answering engine.

    MIT
    Setup
    1 of 5
    Stars
    37.9k
    Release
    10 April 2026
  35. 18 Slowing
    Khoj

    Your AI second brain. Self-hostable. Get answers from the web or your docs. Build custom agents, schedule automations, do deep research. Turn any online or local LLM…

    AGPL-3.0 relicensed 2023
    Setup
    3 of 5
    Stars
    37.9k
    Release
    26 March 2026
  36. 17 Active
    Kotaemon

    An open-source RAG-based tool for chatting with your documents.

    Apache-2.0
    Setup
    Docker
    Stars
    25.8k
    Release
    31 May 2026

Our picks by use case

  • Run models on your own machine

    Ollama downloads and runs open models with one command, and llama.cpp is the C/C++ engine underneath many local AI apps.

  • A ChatGPT-style service for a team

    Open WebUI and LibreChat add accounts, roles and document chat on top of local or hosted models. LibreChat is MIT-licensed; Open WebUI's own license limits changes to its branding.

  • Private chat on a laptop

    Jan and Cherry Studio are desktop apps that run local models or connect to hosted providers, with no server to maintain.

  • Serving models to many users

    vLLM delivers high throughput on GPU servers, and LocalAI exposes text, image and speech models behind OpenAI-compatible APIs.

  • Answers from your documents

    RAGFlow parses complex PDFs and scans, AnythingLLM bundles document chat with a built-in vector database, and Onyx searches company apps while respecting their permissions.

  • One gateway for many providers

    LiteLLM puts more than 100 model APIs behind one OpenAI-style endpoint, with virtual keys, budgets and logs.

    • LiteLLMMITpaid features in repohealth 83

  • Personal agents in your chat apps

    OpenClaw answers in WhatsApp, Telegram, Slack and other channels from your own computer, and Hermes Agent learns reusable skills as it works.

  • Deep research with citations

    GPT Researcher and Local Deep Research plan searches, read sources and write cited reports; the second can run entirely on local hardware.

How to choose an AI chat or local LLM tool

This category mixes three layers that work together. Model runners such as Ollama, llama.cpp and vLLM load a model and serve an API; chat interfaces such as Open WebUI, LibreChat or Jan give people a place to talk to it; RAG platforms add your documents, so answers can cite your own files. Many setups combine one tool from each layer.

Where the model runs

Local models keep prompts on your hardware but need memory: small models run on a laptop, while larger ones need a GPU with plenty of VRAM or several machines joined with exo. Hosted models are stronger and cost per token. Most interfaces here work with both, so you can start hosted and move models local later.

Who will use it

For one person, a desktop app like Jan is enough. A team needs accounts, roles and single sign-on, which LibreChat and Open WebUI provide. For a whole company with knowledge spread across SaaS tools, Onyx adds search that follows each tool's permissions.

Documents and retrieval

If answers must come from your files, compare how each tool reads documents. Simple text splitting works for notes and Markdown; scanned PDFs, tables and slides need layout analysis like the one in RAGFlow. Test candidates with your hardest documents before you decide.

License

Several popular projects here are not OSI open source. Open WebUI forbids removing its branding in larger deployments, LobeHub requires a commercial license for derivative works, and FastGPT restricts multi-tenant SaaS use. If you plan to rebrand or resell, start from a project under a permissive license.

Running costs

A chat interface is light, but models are not. Budget for GPU memory or API spending first, then for the stack around it: LibreChat runs about six services and LobeHub about ten, while AnythingLLM and LocalAI start from a single container.

Drafted with AI from the projects’ READMEs and our data, then checked against that data on 30 October 2026. Report a correction

Which of these are really open source?

Fully open source (30)OSI-approved license, no paid parts in the repository
OpenClaw, Hermes Agent, Ollama, llama.cpp, vLLM, NextChat, AnythingLLM, PrivateGPT, Cherry Studio, LocalAI, exo, Text Generation WebUI, LibreChat, Jan, Chatbox, Open Notebook, Vane, Khoj, SillyTavern, GPT Researcher, llamafile, Kotaemon, MaxKB, DocsGPT, Leon, KoboldCpp, Local Deep Research, Page Assist, big-AGI, Harbor
Open source with paid parts (3)OSI license, but some features are under commercial terms
RAGFlow, LiteLLM, Onyx
Not OSI open source (3)Code is public, use is restricted
Open WebUI (Open WebUI License), LobeHub (LobeHub Community License), FastGPT (FastGPT Open Source License)
License changesNewest first
exo: GPL-3.0 → Apache-2.0, 2025; Jan: AGPL-3.0 → Apache-2.0, 2025; Open WebUI: BSD → Open WebUI License, 2025; GPT Researcher: MIT → Apache-2.0, 2024; LobeHub: MIT → LobeHub Community License, 2024; Chatbox: MIT → GPL-3.0, 2024; Khoj: GPL-3.0 → AGPL-3.0, 2023

Read automatically from each repository’s license files every day. How we read licenses

Questions about open-source AI chat & local LLMs

What are the most popular open-source AI chat & local LLMs projects?

By GitHub stars: OpenClaw (394k), Hermes Agent (259k), Ollama (186k). By our health score, the most actively maintained are AnythingLLM (94), RAGFlow (93), KoboldCpp (93).

Which AI chat & local LLMs projects are fully open source?

30 of the 36 use an OSI-approved license with no paid parts in the repository, including OpenClaw, Hermes Agent, Ollama.

Which are the easiest to self-host?

Judged by the services in their compose files, Hermes Agent, Open WebUI, RAGFlow, NextChat need the least setup.

Which projects have changed their license?

Open WebUI (to Open WebUI License in 2025), LobeHub (to LobeHub Community License in 2024), exo (to Apache-2.0 in 2025), Jan (to Apache-2.0 in 2025). Each project page links to the commit that made the change.

Other categories