OpenAI announced on June 11, 2026 that it will acquire Ona, officially Gitpod GmbH, a German startup whose platform runs AI agents in cloud-based sandboxes, SiliconANGLE reported. Terms were not disclosed. The point of the deal is specific and revealing about where AI coding is heading: it lets OpenAI's Codex assistant keep working on a task for hours or even days, unattended, after the developer has closed their laptop. For software teams, that shifts the coding agent from a tool you supervise keystroke by keystroke to a worker you hand a job and check on later.
What Ona Does: Sandboxes That Outlive Your Workstation
Ona, formerly Gitpod, provides cloud-based sandboxes, isolated, disposable environments where code can run safely, and orchestrates AI agents inside them. The key property is that those sandboxes remain online even when the developer who started them shuts down their own machine, as InfoWorld explained. On a normal setup, an agent running on your laptop dies the moment you close the lid. In Ona's cloud, the agent's work continues, and it draws on far more computing resources than a personal workstation can offer, so long jobs finish faster and without interruption.
Why This Strengthens Codex: From Minutes To Days
This is the engineering reason OpenAI wanted Ona. Codex, OpenAI's software-development assistant, becomes far more useful if it can take on tasks that span hours or days, refactoring a large codebase, running and fixing a long test suite, working through a backlog, without being severed every time the human logs off. Workstation-independent, cloud-resident execution is the infrastructure that makes long-running autonomous agents practical rather than a demo. The technical challenge such systems must solve is keeping a long-running agent on track: an agent left alone for hours can drift, loop, or rack up costs, which is why orchestration, monitoring and sandboxing, exactly Ona's specialty, matter as much as the underlying model.
The Competitive Subtext: OpenAI Versus Anthropic For Developers
The acquisition is also a strategic move. Anthropic has made notable inroads with developers and enterprises, and long-running autonomous agents are the next battleground. By buying the infrastructure to run agents reliably in the cloud, and in a customer's own secure environment, OpenAI is fighting to keep Codex competitive where serious engineering work happens. Ona's entire team, led by CEO and co-founder Johannes Landgraf, will join OpenAI's Codex group once the transaction closes, which remains subject to customary regulatory approvals.
Read more: GitHub Copilot CLI Adds Pre-Commit Security Scanner: LLM Inference at the Detection Layer
What It Means For Software Jobs
Agents that work for hours unattended inevitably raise the question of what happens to software engineering jobs. The honest read is more nuanced than "AI replaces coders." Humans still define the task, set the constraints, review the output, and own the consequences when an agent ships a bug. What changes is the unit of delegation: instead of asking an assistant to complete a function, a developer can hand off a multi-hour chore and spend their time on design and review. That raises productivity, but it also concentrates value in the people who can specify work precisely and judge results, skills that become more important, not less, as the agents get more autonomous.
Frequently Asked Questions
What is Ona and what does it do?
Ona, formerly Gitpod, is a German startup that runs AI agents in cloud-based sandboxes. Its key feature is that those environments stay online after a developer shuts down their own computer, so an agent's work continues uninterrupted.
Why did OpenAI buy Ona?
To strengthen Codex, its software-development assistant. Ona's cloud sandboxes let Codex agents run long tasks, lasting hours or days, without being cut off when a developer logs off, and with more compute than a laptop offers.
Does this mean AI will replace software engineers?
Not directly. Humans still define tasks, set limits, and review and own the output. The change is that developers can delegate longer, unattended jobs, which raises the value of people who can specify and judge work precisely.
ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission.
