Skip to content
Learn · Guide · All Hands AI

OpenHands

An open-source autonomous software-development agent from All Hands AI: it runs in a sandboxed environment with a terminal, editor, and browser, and you can self-host it or use the hosted cloud.

All Hands AI6 min readwww.all-hands.dev

What is OpenHands?

OpenHands is an open-source autonomous coding agent from All Hands AI (formerly the OpenDevin project). It runs in a sandboxed environment with a terminal, code editor, and browser, so it can write code, run commands, and browse the web to complete a software task.

Because it is open source, you control it: self-host the agent, point it at the model of your choice (open or closed), and inspect exactly what it does. All Hands AI also offers a hosted cloud version for teams that do not want to run it themselves.

OpenHands suits engineers who want a transparent, controllable agent and the freedom to choose the model behind it, including self-hosted open weights for data-sensitive work. As an autonomous agent, its output is a draft you review before merging.

Strengths

What it's best for

  • Engineers who want an open-source, self-hostable coding agent they control.
  • Pairing an agent with the model of your choice, including open weights.
  • Data-sensitive work where the agent and model run in your own environment.
  • Teams that want transparency into exactly what the agent does.
Limits

Where it falls short

  • Non-developers wanting a polished, no-setup consumer product.
  • Teams unwilling to run or configure infrastructure (though a cloud option exists).
  • Merging autonomous changes without human review.
How to use it

Ways in

Run OpenHands yourself (commonly via a container) and connect the model you want, or use the hosted OpenHands Cloud to skip setup.

Give it a task and a repository; it works in a sandbox with a terminal, editor, and browser, and proposes the changes for you to review.

How to use it

Getting the most out of it

Choose the model to fit the job and your constraints; a capable coding model such as Devstral or a frontier model both work behind the same agent.

Keep tasks scoped and review the diff. For sensitive code, run both the agent and the model in your own environment.

Pricing

What OpenHands costs

Approximate, in USD, as of June 2026. Prices change often. Confirm on the official site before you rely on them.

Open source

$0 (self-host)

The agent is free and open source; you pay only your own compute and any model API you choose. Confirm the license on the official site.

OpenHands Cloud

Usage-based (credits)

A hosted version with credits for teams that do not want to self-host. Confirm current pricing on the official site.

Visit the official OpenHands site
Try it

Example prompts

Copy these into OpenHands as starting points, then adapt them to your task.

Implement a taskCopy prompt
Here is the repository and the task. Implement it, run the tests, and show me the diff and the test output before you finish.
Investigate a bugCopy prompt
Reproduce this bug from the steps, find the root cause by reading the code and running it, and propose a fix.
Choose a modelCopy prompt
Configure the agent to use a self-hosted open-weight model, and confirm no code leaves my environment during a run.
FAQ

OpenHands
common questions.

Direct answers to the questions we get asked the most. If yours isn't covered, write to the team.

Work with SDEN

Putting AI into production?

We help teams choose the right models and ship them securely, self-hosted when data demands it. And we hand you the keys to run them in-house.

OpenHands guide · SDEN