AI Agents
open-multi-agent
Open Multi Agent — a self-hosted TypeScript agent runtime with durable approvals and verifiable run records.
★ 7.0k starsMIT
View guide →// agents · EverMind-AI/Raven
Raven
Raven is an open-source ai agents project by EverMind-AI (EverMind-AI/Raven on GitHub). It has 5,303 GitHub stars and 152 forks, is released under the Apache-2.0 license, and was last updated on GitHub on 8 October 2026.
The harness of harnesses: a persistent, self-evolving multi-agent ecosystem built for RSI.
★ 5.3k stars⑂ 152 forksApache-2.0updated 2026-10
▦ visual — this project's README ships no embeddable screenshot,
so here's a placeholder. The guide below is complete regardless.
so here's a placeholder. The guide below is complete regardless.
// project facts
Project facts
- Repository
- EverMind-AI/Raven
- GitHub stars
- 5,303
- Forks
- 152
- License
- Apache-2.0
- Last updated on GitHub
- 8 October 2026
- Category
- AI Agents
- Listing
- Full install guide
- Official site
- Official site
// how it works
How it works
- Clarify the outcome you want the ecosystem to own
Pick a substantial, ongoing objective — a product launch, a codebase, a research program — rather than a one-off question.
- Install Raven
Use the one-line installer for your platform, or the container route if you prefer nothing on the host but Git and Docker.
- Set up your agents and models
Configure the built-in agents and connect any third-party agents you already use.
- Launch and delegate
Start the WebUI with raven web and give Raven its first task from the composer, with skills, playbooks, knowledge and memory attached.
// features
Why people love it
Built-in agent team
Ships Raven-Research, Raven-Code, Raven-Design and Raven-Oncall agents covering research, engineering, design and operations work.
Runtime self-evolution
Agents improve through a runtime self-evolution loop — the project's RSI focus — with persistent memory across tasks.
Connects third-party agents
External agents plug into the same ecosystem, so existing agents join the shared memory and playbooks instead of being replaced.
WebUI and TUI
A browser WebUI (with a task composer exposing skills, playbooks, knowledge and memory) plus a terminal interface.
Proven on full projects
The showcase documents complete end-to-end work: a game development project, an RSI run and a full product launch executed by Raven.
Multiple install paths
One-line script installs, a container route needing only Git and Docker, and editable source-checkout installs for development.
Advertisement
// run it
Run it yourself
- Clone the repository
Clone Raven if you want a source checkout; most users use the installer in the next step.
git clone https://github.com/EverMind-AI/Raven.git cd Raven - Install Raven
Run the official installer on Linux, macOS or WSL2 (a PowerShell route is documented for Windows).
curl -fsSL https://raven.evermind.ai/install.sh | bash - Launch the WebUI
Start Raven's web interface in the background and open it in your browser.
raven web # Opens the WebUI; stop it later with: raven web --stop
// use cases
What you can build
- Running a complete product-launch project with a persistent agent team
- Letting specialized agents handle research, code and design in one shared memory
- Studying recursive self-improvement patterns in a real multi-agent runtime
- Coordinating your existing third-party agents under one harness
Advertisement
// faq
Questions, answered
Who maintains Raven?
Raven is maintained by EverMind-AI, part of the EverMind ecosystem.
Is Raven free?
Yes — it is open source under the Apache-2.0 license. You pay for the models your agents call.
Does it need a paid API key?
Raven's agents drive LLMs, so you need model access through the providers you configure; Raven itself charges nothing.
How does it compare to AutoGen or CrewAI?
Those are frameworks for assembling agent teams in code; Raven is a persistent, productized ecosystem — built-in agents, shared memory, a WebUI, and a self-evolution loop — that also connects external agents.
How do I get started?
Run the one-line installer, then raven web, and delegate your first task from the composer.
// keep exploring
Similar tools
AI Agents
moltis
Secure persistent personal agent server in one Rust binary — sandboxed, multi-provider, on your hardware
★ 2.9k starsMIT
View guide →AI Agents
OpenViking
A self-evolving context database unifying agent memory, RAG, and skills.
★ 39.4k starsAGPL-3.0
View guide →AI Agents
ruflo
Agent harness for multi-agent swarms: deploy, coordinate and federate autonomous workflows
★ 74.1k starsMIT
View guide →AI Agents
EverOS
One portable, local-first memory layer for every AI agent - Markdown-native, user-owned and self-evolving.
★ 13.4k starsApache-2.0
View guide →AI Agents
reef
Reef — infrastructure for continually self-improving agents: weight-training and harness-evolving deployments.
★ 7.5k starsApache-2.0
View guide →// alternatives
Alternatives in AI Agents
The most-starred other AI Agents projects in the LibreAI catalogue — the realistic shortlist if Raven is not the right fit.
- openclaw★ 391.6k
A personal AI that really does things — on any OS, with your data staying yours.
- superpowers★ 294.9k
An agentic skills framework & software development methodology that works.
- ECC★ 275.2k
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
- hermes-agent★ 252.0k
The agent that grows with you — a Python framework for adaptive AI agents.
- deepseek-harness★ 245.5k
DeepSeek's plugin-based agent harness where everything is a plugin.
// more in agents
More in AI Agents
Nearby AI Agents projects in the catalogue, closest in stars to Raven — keep exploring the shelf either side of it.
- TaskingAI★ 5.4k
An open-source platform for AI-native application development — assistants, tools, RAG and models in one API.
- zotero-mcp★ 5.3k
Connect your Zotero research library to Claude and other AI assistants through MCP.
- eve★ 5.5k
Vercel's open framework for building agents that run anywhere.
- FrontierAgent★ 5.2k
An open agent framework with a native terminal TUI, ReAct and Agent Team modes — one command to run.
- ruoyi-ai★ 5.7k
Enterprise-grade AI agent framework with multi-provider LLM management, secure knowledge bases and high-precision RAG, visual workflow orchestration, and mul…
- 01★ 5.2k
Talk to your computer — an open-source voice interface.
Get Raven
Free and open — grab the source, star the repo, and start creating.