Awesome Self-Hosted AI

Awesome Self-Hosted AI

Lists 309 open-source AI projects you can run on your own hardware or server, grouped into sections such as LLM runtimes, agents, RAG and speech, with a weekly health check.

Made by glincker on GitHub. These notes are Mindket's, written from its README; the project's own files and license are in its repository.

About

What it is: Awesome Self-Hosted AI is a curated list of open-source AI projects meant for your own machines. They run on hardware or servers that you control. It holds 309 projects, grouped into sections such as LLM runtimes, agents, RAG, speech and voice agents. Every week an automated job looks each project up through the GitHub API.

Who it suits: It suits people who want to host AI tools themselves and need one place to compare options. Developers, small IT teams and research groups can all find a starting point here.

Industry | Who uses it | How they use it | Example
Software development | Developers who want local AI tools | Find a coding assistant or an LLM runtime to run themselves | Browsing the Coding Assistants section
IT and operations | Teams that host their own tools | Compare chat interfaces, model gateways and observability tools | Choosing a self-hosted chat app for a team
Data and research | Research and data teams | Look through RAG, vector database and evaluation projects | Picking a vector database to test

You can skip it if you want one ready-made app, since this is a directory of other projects and not an installer.

Where to use it:

• GitHub: the list is a README with a link to each project's own repository
• Your own hardware or server: where the listed projects run
• Docker: some entries, such as Sim, can be self-hosted with it
• A live health table beside the list, kept in a file called LIVE.md

Use case | Who | What they get back
Find an agent framework | Developers | Entries such as CrewAI, LangGraph and Pydantic AI, each with a line on what it does
Pick a chat front end | Individuals and teams | Entries such as AnythingLLM, Chatbox and Cherry Studio
Check a project's health | Anyone choosing | Stars, last push and license for each project in the live health table

How to use it: Open the section that matches your need and read the one-line entries. Follow a link to reach the project's own repository and its setup steps. Use the live health table to compare projects, and look in the unmaintained file for ones that dropped out.

Tool | What it does
LIVE.md | The live health table with stars, last push and license for each project
UNMAINTAINED.md | Where projects go when they are archived, deleted or have had no activity for over a year

Example: Say you want an agent framework you can run yourself. Open the Agents and Automation section, which has 27 entries. CrewAI organizes role-based teams of agents in your Python environment. LangGraph builds stateful agent workflows as graphs. Pydantic AI is a Python agent framework built around type safety. Then open the health table to compare them.

For newcomers and experts: If you are new, open the Chat Interfaces section first. It has 11 entries, including AnythingLLM and Chatbox, and each one is a short description with a link to the project.

If you know the field, the upkeep is what sets the list apart. A weekly job checks every entry through the GitHub API. Entries that are archived, deleted or idle for more than a year are sent to UNMAINTAINED.md, which keeps the main list current. Rising and Source-Available are separate sections with 25 entries each.

Projects are listed by section
Weekly check through the GitHub API
Active projects stay in the list
Stale ones move to UNMAINTAINED.md

Good to know:

• It is a list, not an installer, so each project has its own repository, setup and license
• When the list was refreshed on 2026-11-02, 308 of its 309 projects had recent activity within 90 days and none was marked unmaintained
• Each entry is a short summary, so read the project's own repository before you rely on it

Title: Projects in six of the larger sections
LLM Runtimes and Inference: 38 projects
Agents and Automation: 27 projects
Coding Assistants: 25 projects
RAG and Vector Databases: 24 projects
Observability and Evals: 20 projects
MCP Servers and Gateways: 19 projects

On GitHub, 7 November 2026

• Stars: 113
• Forks: 15
• Open issues and pull requests: 3
• Language: JavaScript
• License: CC0-1.0
• Topics: ai-agents, awesome, awesome-list, homelab, llm, local-ai, local-llm, mcp-server, ollama, open-source, privacy, rag

Links

• Code: https://github.com/glincker/awesome-selfhosted-ai
• Website: https://github.com/glincker/awesome-selfhosted-ai/blob/main/LIVE.md
• Releases: https://github.com/glincker/awesome-selfhosted-ai/releases
• Issues: https://github.com/glincker/awesome-selfhosted-ai/issues
• Star history: https://www.star-history.com/#glincker/awesome-selfhosted-ai&Date

Get the code

git clone https://github.com/glincker/awesome-selfhosted-ai.git

Then follow the setup in its README. Nothing in this kit runs by itself.