Updated

AutoSwarm is an open-source multi-agent coordinator from the HKU Data Intelligence Lab (HKUDS) that turns the coding and general agents already installed on your machine into one coordinated team.

Pricing Free open source
Visit site

Follow AutoSwarm to get an email when its listing changes, such as new pricing or features. One digest a day at most, and you can unsubscribe anytime.

Company HKUDS
Pricing model Open Source
Paid plans None listed
Category peers 37
01Overview

It installs as a skill into the agent you already use, then selects agents by capability, shares context and memory between them, runs subtasks in parallel and verifies and synthesizes the results.

02Key features
  1. 01

    Coordinates Codex, Claude Code, OpenCode, Cursor, GitHub Copilot CLI, Muse Code, CodeBuddy, Hermes, nanobot, DeepSeek Harness and Grok Build

  2. 02

    Installs as a skill/plug-in so any supported agent can be the entry point (`autoswarm setup --host <agent>`)

  3. 03

    Capability-based agent selection with shared context and memory across agents

  4. 04

    Parallel subtask execution with result verification and synthesis

  5. 05

    `autoswarm doctor --probe` checks that each provider actually responds and can resume natively

  6. 06

    Reuses existing models, credentials and proxy settings on Windows, macOS and Linux

03Pricing

Open Source

Free and open source under the MIT license; you bring your own installed agent CLIs and pay their subscription or API costs.

04Who it's for

Best for

Developers who already run several coding agents and want one task fanned out across them without manually switching interfaces and copying context.

Not ideal for

Users with a single agent or no local agent CLIs, and teams that need a mature, stable orchestration product; it is a research-led project with no tagged releases yet.

05Verdict
A promising, research-backed way to make your existing agents work as one team, best suited to power users comfortable with an early-stage tool.
ai.dosa.dev editorial · Oct 27, 2026
06Recent updates

Released publicly in 2026 with the accompanying arXiv paper "AutoSwarm: Automating Agent Swarms" (2610.33048); trending on GitHub in late October 2026 with roughly 2,000 stars and daily commits.

As of

07Pricing & FAQ
Is AutoSwarm free?

Yes - AutoSwarm is Open Source. Free and open source under the MIT license; you bring your own installed agent CLIs and pay their subscription or API costs.

Is AutoSwarm open source?

Yes - AutoSwarm is open source. Free and open source under the MIT license; you bring your own installed agent CLIs and pay their subscription or API costs.

Who is AutoSwarm best for?

Developers who already run several coding agents and want one task fanned out across them without manually switching interfaces and copying context.

Who is AutoSwarm not ideal for?

Users with a single agent or no local agent CLIs, and teams that need a mature, stable orchestration product; it is a research-led project with no tagged releases yet.

What are the best AutoSwarm alternatives?

The closest AutoSwarm alternatives on ai.dosa.dev are Jules, OpenHands, SWE-agent - all listed under Autonomous & Async Agents.

Who makes AutoSwarm?

AutoSwarm is developed by HKUDS. It is listed in the Autonomous & Async Agents category on ai.dosa.dev.

08More in Async Agents
09Tags
  • Autonomous Agents
  • Multi-Agent
  • Orchestration
  • Open Source
  • Agent Skills
Track AutoSwarm in your AI stack

Favorite this tool to revisit it later, or Zap it to contribute to the public vote count.

Compare Async Agents Sign in to save
Content on this page is AI-generated. Please verify details with the vendor's website for accuracy. Are you the maker? Get your Featured badge ← Back to directory