OpenHuman: An open-source agent harness with a Rust core (42k stars)
What it does
All text below is quoted directly from the project README.
>An open-source agent harness with a Rust core: lightweight, modular, and pluggable into whatever LLM, memory, or search engine you already run.
>Early Beta: Under active development. Expect rough edges.
What is OpenHuman?
OpenHuman is a Rust core with a desktop app, a browser UI, a terminal client, and a Rust library wrapped around it. The same core runs all four.
Lightweight and fast
The core runs in-process, not as a separate daemon the UI talks to over a socket.
A cold agent turn takes 102 ms; the full nine-phase bootstrap (config load, registry init, agent build, memory construction, first turn) takes 476 ms.
Pluggable engines
Every engine OpenHuman calls out to is chosen by config, not hardcoded:
- LLM: the managed TinyHumans route, Ollama, LM Studio, MLX, any local OpenAI-compatible server, Claude Code or the Claude Agent SDK, and 26 bring-your-own-key providers including OpenRouter, OpenAI, Anthropic, Google, Groq, Mistral, DeepSeek, Together, and Fireworks.
- Embeddings: the managed Voyage-backed route, or your own Voyage, OpenAI, Cohere, Ollama, or OpenAI-compatible endpoint.
- Web search: managed search included with a subscription, or your own key for Parallel, Brave, Querit, Exa, Tavily, or a self-hosted SearXNG instance.
Workflows
Workflows are saved, typed automation graphs, built on the open-source tinyflows engine. The catalog has 22 node kinds (agent calls, HTTP requests, code, conditions, loops, sub-workflows, approvals, and more), and a graph can trigger on a schedule, an app event, or manually, and can resume mid-run after a pause.
The difference from n8n or Zapier: you describe what you want, the agent drafts the graph, and you review and save it rather than wiring nodes by hand.
Desktop, browser, and terminal
The same core ships three ways: a Tauri v2 and Wry desktop app for Windows, macOS, and Linux; the identical SPA running in any browser (pnpm dev:app:web); and a ratatui-based terminal client (crates/openhuman-tui).
>Download installers from tinyhumans.ai/openhuman or from the GitHub Releases page.
Repository
tinyhumansai/openhuman
Repository stats
|Metric|Value|
|:-|:-|
|Stars|41,507|
|Forks|4,088|
|Last release|v0.64.10 (2026-09-30)|
|Languages|Rust 51.5%, TypeScript 43.1%, JavaScript 3.4%|
|Contributors|191|
|AI contributor accounts (Claude / Codex / Cursor / Copilot)|Claude: 629, Cursor: 29, Codex: 1, Copilot: 1|
|AI co-authored commits|Claude: 649, Cursor: 29, Codex: 1, Copilot: 3|
Main contributor
|Field|Value|
|:-|:-|
|Name|Steven Enamakel|
|GitHub|@senamakel|
|Commits to this repo|19,205 (top contributor)|
|Bio|Building Open Source AGI. "When you're going to change the world, don't ask for permission."|
|Company / Location|@tinyhumansai / Nauvis|
|Website|enamakel.com|
|X / Twitter|@senamakel|
|Followers|419|
|Public repos|89|
|On GitHub since|2017-08-14|
Get the best repos from this community twice a week,100% free: join the newsletter
102 points · 5 comments
Mentions
500 AI Agents on a $10 Server? OpenHuman, Reviewed Can an open-source agent harness really run 500 AI agents on a $10 server? OpenHuman is a Rust agent harness (desktop app + library) that hit #1 on GitHub trending for nine days. Here's what it does, the team's side-by-side test against Hermes, and my verdict. Repo: tinyhumansai/openhuman Benchmarks: tinyhumansai/openhuman-benchmarks Note: the speed, token and memory numbers come from the OpenHuman team's own public benchmarks. Early beta, GPL-3.0 licensed. Try it today, or maybe later? Tell me in the comments. Follow for the next repo. Music: "Inspired" by Kevin MacLeod (incompetech.com) Licensed under Creative Commons: By Attribution 4.0 License creativecommons.org #shorts #aiagents #opensource #rust #github
🔗 openhuman — tinyhumansai/openhuman 🔗 ECC — affaan-m/ECC 🔗 headroom — headroomlabs-ai/headroom ⚡ Quick start: curl -fsSL raw.githubusercontent.com | bash | npx ecc-universal@2.2.3 setup | headroom wrap claude 🌐 All tools + commands: limitx-hub.vercel.app 🚀 Daily AI dispatch in your inbox: hebis-newsletter.beehiiv.com
500 AI Agents, $10 Server?! 3 Trending Repos #Shorts 3 open-source AI agent repos exploding on GitHub today: openhuman (Rust harness, 19.8s/task), ECC (71 agents, 311 skills), headroom (up to 57% token savings). All benchmarks vendor-reported by the repos — independent tests pending. Commands on screen; all links below. 🔗 tinyhumansai/openhuman 🔗 affaan-m/ECC 🔗 headroomlabs-ai/headroom ⚡ curl -fsSL raw.githubusercontent.com | bash ⚡ npx ecc-universal@2.2.3 setup ⚡ headroom wrap claude 🌐 All tools + commands: limitx-hub.vercel.app 🚀 Daily AI dispatch: hebis-newsletter.beehiiv.com #Shorts #AI #AIAgents #OpenSource #GitHub #Rust #Coding #TechHindi #LocalLLM #MachineLearning
🚀 [Fastest Growing GitHub Repos] openhuman: The Free Tool Taking Over GitHub! #shorts Discover tinyhumansai/openhuman: The fastest, cheapest, most efficient open-source agent harness. Run more than 500 agents on a $10 VPS. 💬 Comment "OPENHUMAN" below for the direct link, or grab it from the links below: 👉 Interactive Star History & Live Analysis: githubrepo.cloud ⭐ Embed Live Verified Badges in your README: githubrepo.cloud 🔗 Direct GitHub Repository: tinyhumansai/openhuman 📊 Repository Specs: ⭐ Stars: 41,945 🍴 Forks: 4,126 💻 Language: Rust 📜 License: GPL-3.0 #shorts #openhuman #agentorchestration #aiagents #aiassistant #desktop #ai #artificialintelligence #machinelearning
OpenHuman is the fastest growing opensource agent harness tinyhumansai/openhuman (github updated) t.co
OpenHuman: An open-source agent harness with a Rust core (42k stars) What it does All text below is quoted directly from the project README. >An open-source agent harness with a Rust core: lightweight, modular, and pluggable into whatever LLM, memory, or search engine you already run. >Early Beta: Under active development. Expect rough edges. What is OpenHuman? OpenHuman is a Rust core with a desktop app, a browser UI, a terminal client, and a Rust library wrapped around it. The same core runs all four. Lightweight and fast The core runs in-process, not as a separate daemon the UI talks to over a socket. A cold agent turn takes 102 ms; the full nine-phase bootstrap (config load, registry init, agent build, memory construction, first turn) takes 476 ms. Pluggable engines Every engine OpenHuman calls out to is chosen by config, not hardcoded: - LLM: the managed TinyHumans route, Ollama, LM Studio, MLX, any local OpenAI-compatible server, Claude Code or the Claude Agent SDK, and 26 bring-your-own-key providers including OpenRouter, OpenAI, Anthropic, Google, Groq, Mistral, DeepSeek, Together, and Fireworks. - Embeddings: the managed Voyage-backed route, or your own Voyage, OpenAI, Cohere, Ollama, or OpenAI-compatible endpoint. - Web search: managed search included with a subscription, or your own key for Parallel, Brave, Querit, Exa, Tavily, or a self-hosted SearXNG instance. Workflows Workflows are saved, typed automation graphs, built on the open-source tinyflows engine. The catalog has 22 node kinds (agent calls, HTTP requests, code, conditions, loops, sub-workflows, approvals, and more), and a graph can trigger on a schedule, an app event, or manually, and can resume mid-run after a pause. The difference from n8n or Zapier: you describe what you want, the agent drafts the graph, and you review and save it rather than wiring nodes by hand. Desktop, browser, and terminal The same core ships three ways: a Tauri v2 and Wry desktop app for Windows, macOS, and Linux; the identical SPA running in any browser (pnpm dev:app:web); and a ratatui-based terminal client (crates/openhuman-tui). >Download installers from tinyhumans.ai/openhuman or from the GitHub Releases page. Repository tinyhumansai/openhuman Repository stats |Metric|Value| |:-|:-| |Stars|41,507| |Forks|4,088| |Last release|v0.64.10 (2026-09-30)| |Languages|Rust 51.5%, TypeScript 43.1%, JavaScript 3.4%| |Contributors|191| |AI contributor accounts (Claude / Codex / Cursor / Copilot)|Claude: 629, Cursor: 29, Codex: 1, Copilot: 1| |AI co-authored commits|Claude: 649, Cursor: 29, Codex: 1, Copilot: 3| Main contributor |Field|Value| |:-|:-| |Name|Steven Enamakel| |GitHub|@senamakel| |Commits to this repo|19,205 (top contributor)| |Bio|Building Open Source AGI. "When you're going to change the world, don't ask for permission."| |Company / Location|@tinyhumansai / Nauvis| |Website|enamakel.com| |X / Twitter|@senamakel| |Followers|419| |Public repos|89| |On GitHub since|2017-08-14| Get the best repos from this community twice a week,100% free: join the newsletter
GIthub orgs cant claim Sending to the founder listed on the org github.com/senamakel x.com t.co
THIS IS 100x BETTER THEN OPENCLAW OR HERMES #1 trending in GitHub for weeks openhuman sending fees to the GitHub & contracting the owner. tinyhumansai/openhuman
this open-source agent harness is currently the #1 trending tool on GitHub. it’s built to be lighter than OpenClaw and Hermes, using less RAM + CPU while running faster. if you care about efficiency, this one’s worth checking out. tinyhumansai/openhuman
this open-source agent harness is the #1 trending tool on GitHub right now. that’s because it beat OpenClaw and Hermes on efficiency. it’s built to be more lightweight, uses less RAM & CPU, and delivers results much faster. 🔗: tinyhumansai/openhuman t.co