OpenClaw
Self-Hosted AIThe AI that actually does things
Best Beginner-Friendly Self-Hosted AI for self-hosting
The 9 best Beginner-Friendly self-hosted ai you can self-host, ranked by community traction.
Hermes Agent
Self-Hosted AIThe AI agent that grows with you
OpenCode
Self-Hosted AIThe open source AI coding agent
Replaces Claude Code, Cursor
PicoClaw
Self-Hosted AIUltra-lightweight personal AI assistant
LibreTranslate
Self-Hosted AIFree and Open Source machine translation API
Replaces Google Translate, DeepL
Spacebot
Self-Hosted AIAI agent for teams, communities, and multi-user environments
ChatGPT UI
Self-Hosted AIA ChatGPT web client that supports multiple users, multiple languages,
Chatd
Self-Hosted AIChat with your documents using local AI
Replaces NotebookLM
OpenReader
Self-Hosted AIA document reader with Text-to-Speech capabilities
No apps match these filters.
At a glance
| App | Deploy effort | Difficulty | Min. RAM | License | ★ Stars |
|---|---|---|---|---|---|
| OpenClaw | ≈5-minute setup | Easy | 512 MB | free | 391,924 |
| Hermes Agent | ≈5-minute setup | Easy | 512 MB | free | 253,548 |
| OpenCode | ≈5-minute setup | Easy | 512 MB | free | 213,299 |
| PicoClaw | ≈5-minute setup | Easy | 512 MB | free | 30,020 |
| LibreTranslate | ≈5-minute setup | Easy | 512 MB | free | 17,069 |
| Spacebot | ≈5-minute setup | Easy | 512 MB | free | 2,413 |
| ChatGPT UI | ≈5-minute setup | Easy | 512 MB | free | 1,625 |
| Chatd | Read-the-docs project | Easy | 8,192 MB | MIT | 1,079 |
| OpenReader | ≈5-minute setup | Easy | 512 MB | free | 540 |
Quick picks
- Most adopted: OpenClaw (391,924 GitHub stars).
- Lightest footprint: OpenClaw (from 512 MB RAM, against 8,192 MB for the heaviest here).
- Fastest to deploy: OpenClaw (one-click platform packages).
- Licensing: 8 of 9 use free; check the rest individually if you plan to modify and redistribute.
Every option here is open-source and self-hostable, tagged Beginner-Friendly within the self-hosted ai category. Deploy effort is graded from published container recipes, not vendor claims — see the methodology.