DigitalOcean Adds One-Click Deployment for Y Combinator’s QM
DigitalOcean now offers what it describes as the first one-click deployment for QM, Y Combinator’s open-source agent harness, through the DigitalOcean Marketplace. The setup is designed to give startups QM’s core services, sandboxes, inference, data and storage in a single deployment.
- DigitalOcean says QM is now available through a one-click Marketplace deployment.
- QM was built by Y Combinator for Slack and the web, then open-sourced in July.
- The deployment includes compute, sandboxes, inference, PostgreSQL and file storage.
- Teams can choose among harnesses and models including Pi, OpenCode, Codex and Claude Code.
- Managed Agents is currently offered as a public preview.
QM moves from YC teams to a Marketplace deployment
Y Combinator built QM as a multiplayer agent harness for Slack and the web. According to DigitalOcean, the system was used across YC teams including accounting, legal and engineering before Y Combinator open-sourced it in July so that any startup could use a similar setup.
DigitalOcean says the Marketplace deployment is available starting today, November 10, 2026. The company describes it as the first one-click deployment for QM, giving teams a way to start the system without assembling its infrastructure across multiple providers.
What the one-click deployment includes
The DigitalOcean Marketplace option brings QM’s supporting services together on one bill. It sets up a Droplet for QM’s core service, Managed Agents for isolated sandboxes, and Serverless Inference without an extra API key.
The deployment also includes Managed PostgreSQL for sessions, memory and the job queue, along with Spaces for files and snapshots. DigitalOcean says this combination is intended to reduce the setup work required to run QM and give teams a complete infrastructure stack from the start.
Joshua France, Head of YC Labs at Y Combinator, said the company open-sourced QM “so companies can own the agents that run their business.” He added that “in one click, a team has QM running with sandboxes, inference, data and storage already in place.”
Teams can choose their harnesses and models
QM is designed to let each team select its harness and model. DigitalOcean lists Pi, OpenCode, Codex and Claude Code as examples, and says QM runs in a range of environments.
DigitalOcean contributed Managed Agents to QM as a built-in sandbox option for users on supported versions. Paddy Srinivasan, CEO of DigitalOcean, said the companies are working “to bring open intelligence to YC founders,” adding that startups can choose their models, harnesses and data while running them on infrastructure they control.
Users can deploy QM from the DigitalOcean Marketplace at do.co/qm and consult the QM documentation. DigitalOcean says Managed Agents is offered as a public preview and does not guarantee that public preview services will perform for production-level workloads. The company also says plans for QM and YC founders may change without notice.
