What Is OpenHands?
OpenHands is an open source platform for AI agents that perform software engineering work. It can be used as a ready made coding agent through graphical and hosted experiences or as a development framework for teams building their own agent based engineering systems.
Agents With Real Development Tools
An OpenHands agent can inspect files execute shell commands interact with development environments and work through repository tasks. Hosted cloud workflows make it easier to start quickly while local deployments give users more control over infrastructure model configuration and data handling.
Build On Top Of The Platform
The Software Agent SDK and API provide a path for developers who need OpenHands capabilities inside custom services automation or research projects. This makes the project relevant both to everyday coding tasks and to engineers experimenting with how software agents should be orchestrated.
Who Should Consider OpenHands?
- Developers who prefer open source coding agents
- Researchers evaluating autonomous software engineering
- Teams that want self hosted agent infrastructure
- Companies connecting agents through an API
- AI engineers building specialized coding agents
Key features
- Open Source Agent Platform: Provides inspectable software engineering agent infrastructure.
- Hosted Cloud: Runs tasks without local environment setup.
- Local Deployment: Supports self managed agent execution.
- Shell And File Tools: Gives agents practical development capabilities.
- Repository Integrations: Connects common source control services.
- Software Agent SDK: Enables custom agent applications.
- REST API: Supports programmatic agent workflows.
- Model Flexibility: Works with supported hosted or self configured models.
Use cases
Autonomous Coding,Agent Research,Self Hosted Coding Agents,Repository Automation,Bug Fixing,Feature Development,Agent API Integration,Software Engineering Experiments
How to use
How To Use OpenHands
- Choose Cloud Or Local: Start with the hosted OpenHands service for convenience or follow the local deployment documentation for more infrastructure control.
- Configure Access: Sign in for cloud use or configure the model and runtime required by your local installation.
- Connect A Repository: Authorize a supported source control service or load the project in the local environment.
- Create A Task: Describe the engineering objective and include acceptance criteria and relevant context.
- Start The Agent: Let OpenHands inspect the repository use development tools and work through the requested task.
- Observe Tool Activity: Review commands file edits and reasoning artifacts while the session progresses.
- Inspect The Result: Check generated code tests and any pull request or patch before accepting it.
- Automate When Ready: Use the SDK or API only after the workflow is reliable enough for repeated or integrated use.
Best for
Developers, Researchers, AI Engineers, Engineering Teams, Agent Builders
Integrations
GitHub,GitLab,Bitbucket,API,SDK,MCP,Local Runtime,Cloud Runtime
Commercial use
Depends On Open Source Components Hosted Service Terms And Model Provider Terms
