Build better software with autonomous cloud agents

Offload toil, ship faster. OpenHands automates real development tasks with transparency, governance, and choice.

How OpenHands works

1. Delegate

Use OpenHands in your existing tools or via custom SDK-powered workflows

2. Execute

Agents run autonomously in secure Docker sandboxes

3. Review

OpenHands delivers clean, well-tested PRs - and can automate your code reviews, too

Asynchronous task execution

Model-agnostic flexibility (Anthropic, OpenAI, Gemini, custom)

Secure sandbox runtime (BYOK, RBAC, audit trails)

Agent SDK for custom workflows

GUI, CLI, or headless operation

Task execution

Local

Local: Open source and runs anywhere

SaaS: Fully managed and ready to scale instantly

Self-Hosted: Deploy in your own cloud or on-prem environment for full control

GUI: Visual workspace

CLI: Terminal-based control

Developer toolchain: Jira, GitHub, GitLab, Bitbucket, Slack and more

Secure, scalable architecture by design

Each agent runs in an ephemeral, containerized environment - isolated, auditable, and policy-governed.

Your own agents

Extend OpenHands with your own agents

Code review agents

Production triage agents

Release note agents

The only open, secure, and scalable agent platform built for real engineering work

Building the open standard for autonomous software development.

OpenHands is the foundation for secure, transparent, model-agnostic coding agents - empowering every software team to build faster with full control.