Self-hosted ChatGPT alternatives

Self-hosted ChatGPT alternatives

Chat interfaces and assistant apps for private AI deployments.

Projects 6
Updated regularly
All projectsSuggest a project

ChatGPT alternatives

ChatGPT-style apps you can run with your own models, data, and infrastructure.

Use this page when you need a complete chat product, not just a model backend. These projects provide the user-facing assistant experience: conversations, file uploads, model selection, user access, and private deployment options.

Interface layer

These tools give users a place to chat, manage conversations, and work with files. They usually connect to a separate model runtime or API.

Different product shapes

Some are simple local chat apps. Others add teams, connectors, agents, knowledge bases, or desktop-first workflows.

Why it works

  • Open WebUI and NextChat for lightweight starts

    Open WebUI is a good Docker-first browser UI around Ollama or OpenAI-compatible endpoints. NextChat is a lighter cross-platform assistant for web, desktop, and mobile-style use.

  • LobeChat and LibreChat for polished chat portals

    LobeChat focuses on a refined chat experience. LibreChat is stronger when users, presets, agents, MCP, and admin controls matter.

  • AnythingLLM, Onyx, and Jan for specific workflows

    AnythingLLM fits document workspaces, Onyx fits team knowledge and connectors, and Jan fits desktop-first offline chat.

Curated repositories

Open source ChatGPT-style apps

6 projects
open-webui

open-webui/open-webui

open-webui

134.3k

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

19.1k|Python
NOASSERTION
ollamaollama-webuillm
danny-avila

danny-avila/LibreChat

danny-avila

36.1k

Enhanced ChatGPT Clone: Features Agents, MCP, DeepSeek, Anthropic, AWS, OpenAI, Responses API, Azure, Groq, o1, GPT-5, Mistral, OpenRouter, Vertex AI, Gemini, Artifacts, AI model switching, message search, Code Interpreter, langchain, DALL-E-3, OpenAPI Actions, Functions, Secure Multi-User Auth, Presets, open-source for self-hosting. Active.

7.4k|TypeScript
MIT
aichatgptclone
Mintplex-Labs

Mintplex-Labs/anything-llm

Mintplex-Labs

59.1k

The all-in-one AI productivity accelerator. On device and privacy first with no annoying setup or configuration.

6.4k|JavaScript
MIT
raglmstudiolocalai
onyx-dot-app

onyx-dot-app/onyx

onyx-dot-app

28.6k

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

3.8k|Python
NOASSERTION
enterprise-searchragai-chat
janhq

janhq/jan

janhq

42.2k

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

2.8k|TypeScript
NOASSERTION
llamacpplocalaiself-hosted
ChatGPTNextWeb

ChatGPTNextWeb/NextChat

ChatGPTNextWeb

87.8k

✨ Light and Fast AI Assistant. Support: Web | iOS | MacOS | Android | Linux | Windows

59.8k|TypeScript
MIT
chatgptnextjsvercel

Selection guide

How to choose a self-hosted ChatGPT alternative

Start with the deployment shape and the user experience you need. A solo local setup has different requirements than a team chat portal with authentication, connectors, and admin controls.

  • For the easiest browser UI

    Open WebUI is the practical default when you want a Docker-friendly interface around Ollama or OpenAI-compatible endpoints. NextChat is lighter when you want a fast cross-platform chat client.

  • For multi-provider chat

    LibreChat is a stronger fit when users need several providers, agents, MCP support, search, and secure multi-user access.

  • For document-heavy work

    AnythingLLM and Onyx fit better when the chat experience needs workspaces, files, connectors, or a private knowledge base.

Backend fit

Chat app vs model runtime

A ChatGPT alternative is the product users interact with. It still needs a model backend, which may be Ollama, LocalAI, vLLM, llama.cpp, an OpenAI-compatible endpoint, or a hosted provider.

  • Use this page for the interface

    Conversation history, file uploads, users, model switching, tools, agents, and knowledge features belong here.

  • Use runtimes for serving models

    Ollama, LocalAI, llama.cpp, and vLLM are important, but they are the inference layer rather than the full chat product.

Team features

What matters for private teams

For team usage, compare authentication, admin controls, data storage, connector support, deployment complexity, and whether the app can run fully inside your environment.

  • LobeChat

    A polished multi-provider chat experience, but self-hosting can require more infrastructure than a simple single-container tool.

  • Onyx

    Best framed as a team knowledge and enterprise-search platform with chat, connectors, RBAC, SSO, and governed access.

Suggested additions

Strong candidates not yet in the registry

Chatbot UI

mckaywrigley/chatbot-ui

9/10

A classic ChatGPT-style web UI with a local/self-hosted path and support for local model workflows. Good candidate once it is added to the upstream registry.

View repository

BionicGPT

bionic-gpt/bionic-gpt

8.8/10

An on-prem ChatGPT replacement for teams, with assistants, RAG, access controls, audit needs, and enterprise deployment concerns.

View repository

Khoj

khoj-ai/khoj

8.6/10

A self-hostable personal AI app with chat, documents, web access, agents, and local or online model support. Broader than a pure ChatGPT clone, but a strong fit for private assistant intent.

View repository

Libre WebUI

libre-webui/libre-webui

7.9/10

A self-hosted AI chat interface with local-first positioning, Docker/Kubernetes deployment, auth, plugins, and RAG. Promising, but newer than the main curated options.

View repository

Related pages

Keep browsing

Local model runtimes and inference servers

Inference servers and local runtimes for serving models on your machine, server, or private cloud.

Self-hosted RAG tools

Knowledge-base apps, retrieval frameworks, and document pipelines for private data and production AI systems.

Vector databases and retrieval storage

Databases and search layers for embeddings, metadata filtering, persistence, and semantic retrieval.

Agents, workflows, and app builders

Agent frameworks, workflow engines, and app builders for repeatable AI-powered processes.

AI developer tools

Coding assistants and repo-aware tools that can run locally or inside private development environments.

FAQ

Questions answered

What is the best self-hosted ChatGPT alternative?

For many local-model setups, start with Open WebUI. NextChat is a lighter cross-platform option. LobeChat is strong for a polished chat experience. LibreChat works well as a multi-provider ChatGPT-style portal. AnythingLLM fits document workspaces, Onyx fits team knowledge and connectors, and Jan fits desktop-first offline chat.

Do self-hosted ChatGPT alternatives include the model?

Usually not. Most apps connect to model backends such as Ollama, LocalAI, vLLM, llama.cpp servers, OpenAI-compatible APIs, or hosted providers. The chat app is the user-facing layer; the runtime is often separate.

Should Ollama be listed as a ChatGPT alternative?

Ollama belongs on the local runtimes page. It becomes a ChatGPT-like experience when paired with a UI such as Open WebUI, LobeChat, LibreChat, AnythingLLM, Jan, or another chat interface.

Which self-hosted ChatGPT alternative is easiest to run with Docker?

Open WebUI is usually the easiest Docker-first starting point for a browser-based private chat UI. NextChat is lighter and can be deployed through web or cross-platform paths. LibreChat, AnythingLLM, Onyx, and LobeChat can be better for teams or specialized workflows, but they may require more services or configuration.

Which self-hosted ChatGPT alternative is best for document chat?

AnythingLLM is a strong fit for document workspaces. Onyx is better when team knowledge, connectors, search, access controls, and enterprise-style governance matter.

Is Jan a self-hosted ChatGPT alternative?

Jan is best understood as a desktop-first, local-first AI app. It fits personal offline chat well, while browser-based server apps are usually a better match for shared team deployments.