OpenHands
An open-source software-development agent platform for working on repositories through local, remote, or cloud execution backends.
By All Hands AI · reviewed 2026-10-21
What OpenHands does
OpenHands provides a browser application, agent runtime, sandboxes, repository workflows, automation, and several agent backends for software engineering tasks. It can connect to hosted models or local servers such as LM Studio, Ollama, vLLM, and SGLang, although capable coding models and an isolated execution environment are important for useful results.
Pricing model
Free open-source self-hosting with optional managed services
Position in the stack
Where OpenHands fits
OpenHands is a software-engineering agent workspace. It combines a user interface, isolated execution backends, repository access, automation, and agent profiles while relying on a separate model.
Typical deployment flow
A practical sequence for evaluating OpenHands before making it part of a permanent stack.
- STEP 1
Choose an execution backend
Run local, remote, or cloud sandboxes with limited repository and credential access.
- STEP 2
Connect a coding model
Use a capable hosted model or local server with enough memory and context.
- STEP 3
Test safely
Measure task completion, diffs, tests, recovery, and permissions on a disposable repository.
Best for
- Repository-scale software engineering tasks
- Running coding agents in isolated workspaces
- Teams comparing local and hosted coding models
Not the right layer for
- Small inline code completions
- Weak local models that cannot reliably use tools
Capabilities
Capabilities refer to the tool's application layer. Hardware and model support can still depend on a connected inference engine.
Deployment and compatibility
- Deployment
- Desktop / localSelf-hostedManaged cloud
- Platforms
- macOS · Windows · Linux · Web
- Hardware backends
- Uses connected model provider
- Install methods
- Docker · Desktop application · Build from source · Managed cloud
Models and integrations
- Model formats
- Uses connected local or hosted model provider
- Common integrations
- GitHubGitLabOllamaLM StudiovLLMSGLangOpenAI-compatible APIs
Strengths
- Full software-engineering workspace
- Local, remote, and cloud execution options
- Supports multiple agent and model backends
Limitations
- Heavier deployment than a terminal-only coding agent
- Local-model quality and context capacity strongly affect success
Planning checklist
Before you choose OpenHands
Answer these questions with the exact models, hardware, users, and data you expect to operate.
- Does the model handle long repository context?
- Where will commands execute?
- How are secrets constrained?
- Do you need shared automation?
OpenHands FAQ
Can OpenHands use local models?
Yes, through LM Studio, Ollama, vLLM, SGLang, and compatible servers.
Is OpenHands only a web UI?
No. It includes agent and sandbox infrastructure behind the interface.
Does it replace Git review?
No. Changes still need normal tests, review, and repository controls.
Official sources
Use these links to confirm current compatibility and installation requirements.
Related tools
A desktop and self-hosted workspace for document chat, RAG, and AI agents.
A node-based workflow engine and interface for local generative image and media models.
An open-source, plugin-first agent harness with local Web UI, tools, sessions, approvals, and extensible model adapters.
Comparisons featuring OpenHands
Use a reviewed comparison when the choice is between two adjacent tools.
Execution evidence
Known working recipes using OpenHands
Recipes connect hardware, a model artifact, tools, settings, verification, and a reportable result.
No verified recipe is linked to this record yet.
Compatibility estimates remain available in the planner. A recipe appears here only after its exact stack and verification protocol are documented.