‍The TLDR: We raised $18.8M in a Series A led by Madrona to build the open, secure, and model-agnostic platform for cloud coding agents. Plus, we're collaborating with AMD to make local agents even better. This funding helps us stick to our mission: powerful coding agents for every developer, for free, forever.

‍

When we started OpenHands, we had a simple but ambitious goal: make sure the future of AI-powered software development happens in the open, not behind closed doors.

Today we can share that we've raised $18.8M in Series A funding led by Madrona, with participation from Menlo Ventures, Pillar VC, Obvious Ventures, Fujitsu Ventures, and Alumni Ventures. We're also announcing a strategic collaboration with AMD to advance agent performance on their hardware.

This isn't just about the money. It's about making sure we can deliver on a promise we made to the developer community: that powerful coding agents will remain open, transparent, and under your control.

The problem we're solving

If you've tried to use AI coding agents at any serious scale, you've probably run into the same frustrations we have.

Open source agents are great for tinkering on your laptop, but try running them across hundreds of repos or with a team of developers, and you'll quickly hit walls around security, access control, and resource management.

Worse, most tools tie you to specific IDEs, specific LLMs, or specific cloud providers. Want to use a different model? Too bad. Need to run on your own infrastructure? Good luck.

When agents are making actual changes to your codebase, you need audit trails, role-based permissions, and the ability to see exactly what happens when something goes wrong.

We built OpenHands to bridge that gap. It's the first platform that gives you the power of cutting-edge coding agents without forcing you to give up control, flexibility, or security.

What makes OpenHands different

Unlike tools that lock you into a single IDE or LLM, OpenHands is open and works with any model. You can run it with Claude, GPT, your own fine-tuned models, whatever works best for your use case.

It scales from one agent on your laptop to thousands running in your cloud infrastructure. Every agent runs in an isolated Docker sandbox for security, and you get organization-level controls like role-based access, audit trails, and quotas.

It integrates with tools you're already using: GitHub, GitLab, Bitbucket, Slack, and Jira. Your team doesn't need to learn entirely new workflows.

What our community has built

The response from the developer community has been incredible. We've hit over 65,000 GitHub stars, 7,000 forks, well over 3 million downloads, and contributions from hundreds of developers worldwide. Engineers at companies like AMD, Apple, Google, Amazon, Netflix, TikTok, NVIDIA, Mastercard, and VMware are using and extending the platform.

But the real validation comes from what early enterprise adopters are achieving:

• 50% reduction in code maintenance backlogs by automating dependency upgrades, lint fixes, and vulnerability remediation
• Vulnerability resolution times cut from days to minutes through automated security sweeps
• Parallel code refactoring across hundreds of repos that would have taken teams weeks to coordinate manually

The software engineering community is increasingly aligning behind OpenHands as the open source standard for doing agentic software development at scale.

Our collaboration with AMD

We're particularly excited about our collaboration with AMD. They're integrating Lemonade Server, an open-source project optimized for AMD hardware, with OpenHands to enable local coding agents that prioritize privacy, cost efficiency, and flexible model selection.

"The integration of Lemonade Server with OpenHands reflects AMD's commitment to open-source innovation and developer choice," said Adrian Macias, Sr. Director of Developer Acceleration at AMD. "This collaboration enables local coding agents that prioritize privacy, cost efficiency, and flexible model selection—while taking advantage of acceleration on Ryzen AI PCs."

This partnership exemplifies what we're trying to build: an ecosystem where developers have real choices about how, where, and with what models they run their coding agents.

How teams are using OpenHands today

Most AI-driven developer tooling focuses on the inner loop of development: it acts as an assistant as the engineer tweaks the code, checks the changes, and iterates.

But OpenHands shines in the outer loop: it can take on whole tasks, and execute them in a sandboxed environment, automating away the entire inner loop for certain categories of tasks, like:

The most common use cases we're seeing fall into a few categories:

Automated maintenance work - Things like dependency upgrades, fixing merge conflicts, and vulnerability sweeps that eat up developer time but don't require human creativity.

Large-scale refactoring - Running coordinated changes across hundreds of repositories in parallel, something that's incredibly tedious to do manually but perfect for agents.

Code quality enforcement - Automated PR reviews, adding unit tests, style fixes, and best practice enforcement that helps teams maintain standards without slowing down development.

The key insight is that agents are best when they handle the repetitive, time-consuming work that prevents developers from focusing on the interesting problems.

What this funding means

This Series A helps us double down on our core mission: putting state-of-the-art coding agents into the hands of every developer, for free, forever.

As Soma Somasegar from Madrona explains: "OpenHands represents a major shift in how software is built. Autonomous agents are transforming from side-projects into core members of the engineering team, and OpenHands' open, model-agnostic approach ensures this transformation happens safely, transparently, and at enterprise scale."

The funding allows us to:

• Keep pushing the research forward - About a third of our engineering team works on advancing the state-of-the-art in coding agents, and all that research goes directly into the open source project
• Scale our infrastructure - Building a platform that can handle thousands of concurrent agents requires serious engineering investment
• Support our growing community - As more developers adopt OpenHands, we need better tooling, documentation, and support systems

Staying true to open source

We've written before about navigating the challenges of commercial open source, and this funding reinforces our commitment to transparency.

Our line in the sand remains the same: the open source project will always be as powerful as possible for individual developers. All our research on code generation goes directly into the open source. Any developer can run state-of-the-art coding agents with whatever LLM they want, without giving their data to anyone.

Commercially, we're building features for teams of developers who need to collaborate with agents at enterprise scale including advanced user management, cloud deployment, and team coordination features.

This approach means we can keep building a public good while creating sustainable revenue to support the project long-term.

What's next

We're just getting started. The future we're working toward is one where every software team has access to powerful AI agents that augment human creativity rather than replace it.

If you're curious about what we're building, check out OpenHands on GitHub. You can run it locally in minutes, or try OpenHands Cloud for free.

And if you want to chat about how OpenHands might fit into your team's workflow. Join our Slack community, we'd love to hear from you.

The future of software development is being written right now. We're making sure it stays open.

‍

Want to stay updated on what we're building? Subscribe to our blog for insights and updates from the OpenHands team.

‍