Hands on the Future

We're at the beginning of a new era in software development. Work is shifting from keystrokes to delegated tasks, from single laptops to coordinated fleets of agents running in environments you control.

This is our point of view on how we build, what we ship, and why it matters… We believe:
‍

The future of software development must be written by engineers
Software development is changing. That change needs to happen in the open, driven by a community of professional developers. That's why OpenHands' software agent is MIT-licensed and trusted by a growing community.

‍

A new class of engineers are emerging
Most engineers write lines of code, maybe with an AI assistant. But Agentic Engineers tackle code maintenance at scale by orchestrating tens to thousands of cloud-based agents. Agent orchestras aren't appropriate for every engineering task, but Agentic Engineers will automate away the toil and busywork weighing down their teams
‍

Scale lives in the cloud
We believe local agents are a bridge, not the destination. Real leverage comes from agents that run in the cloud or your own VPC, so teams can parallelize thousands of tasks and treat automation as infrastructure.
‍

Openness is non-negotiable
AI agents are a transformative technology and an increasingly critical part of the world's infrastructure. We need agents that are open source, transparent, and inspectable if we're going to deploy AI safely and responsibly.
‍

Model choice is a right
We believe you should pick the best model for each job and switch when economics or quality change. This will only become more important as models improve and new tradeoffs emerge. Vendor flexibility is a feature, not a footnote.
‍

Security is a runtime, not a promise
We believe agents must run in isolated, auditable environments with scoped credentials and a clear blast radius. Teams need visibility into how agents are behaving, and the ability to intervene when something breaks

‍

‍

What this means for OpenHands

We are building the open, secure, and model-agnostic platform for cloud coding agents.


Every agent runs in a containerized sandbox you control. Every outcome can be measured and reviewed. You can self-host the open source, or use our cloud. You can bring your own models and switch without drama. You can start local, then scale to thousands of concurrent agents running throughout your codebase.

‍

What this asks of you

Adopt an agentic mindset. Identify toil. Delegate clearly. Measure success. Invest in agent orchestration as a core engineering skill.