Skip to content
News Product launch

OpenAI acquires Ona to give Codex persistent cloud workspaces for long-running agents

· by Pondero Newsdesk

The short version

OpenAI announced on June 11, 2026 that it will acquire Ona, the cloud execution startup behind the Gitpod infrastructure, to extend Codex beyond single-session tasks into multi-hour and multi-day agentic work.

OpenAI acquires Ona to give Codex persistent cloud workspaces for long-running agents

OpenAI announced on June 11, 2026 that it will acquire Ona, a cloud execution and orchestration startup, to extend Codex into work that runs for hours or days rather than finishing within a single session.

What

Ona builds secure, persistent cloud environments where software agents can run continuously with access to an organization's tools, credentials, and systems. The startup describes its sign-in infrastructure via Gitpod and routes its GitHub presence through the gitpod-io organization, confirming Ona's direct lineage from the Gitpod cloud developer environment project. Its CEO, Johannes Landgraf, co-founded and led Gitpod before the company rebranded as Ona to focus on background agents.

Per OpenAI's announcement, more than 5 million people use Codex each week, a figure the company says represents a 400% increase from earlier in 2026. Codex started as a tool for software developers and has expanded to "a wider range of people" doing complex work from an initial request through to a finished result. The acquisition brings Ona's customer-controlled execution model into that ecosystem.

The deal has not yet closed. OpenAI said it is subject to customary closing conditions, including required regulatory approvals. Financial terms were not disclosed. Until closing, the two companies remain separate and independent.

Why it matters

The problem Ona solves is concrete: Codex runs on the machine where a session starts, and when that machine sleeps or the session ends, the work stops. Ona's infrastructure removes that constraint. Agents can continue inside an organization's own cloud environment after a laptop closes, with the company retaining control over where agents run, what they can access, how credentials are scoped, and how activity is logged.

Per OpenAI, the ambition is to let people "delegate more ambitious work without remaining tied to the machine where it began" and to check in from any device rather than monitor a local process.

That positions Codex to compete more directly with platforms that already offer persistent agentic execution. Cursor, GitHub Copilot, and Devin all operate in this space. Ona's own website listed comparison pages for Cursor, Codex, GitHub Copilot, Devin, Claude Code, and Factory before the acquisition announcement. The Ona team will join OpenAI's Codex group after closing to build out enterprise-grade persistent execution.

Context

Ona reported that 2 million developers had worked in its cloud environments before the OpenAI deal. The startup counted BNY, GSR, Vanta, Pearson, and EquipmentShare among its customers. Johannes Landgraf framed the combination this way: "Agents need more than intelligence; they need a trusted workspace. We built Ona to give agents cloud environments with the context, control and collaboration enterprises require. Joining OpenAI lets us bring that foundation into Codex." That is the vendor's own framing, not an independent assessment.

Thibault Sottiaux, Core Products Lead at OpenAI, said the goal is to make Codex "easier to deploy securely across production workflows for customers operating at the highest standards of trust and scale." Again, vendor-attributed.

What to watch next

Three things are worth tracking. First, the regulatory review timeline for a US company acquiring what remains a German-incorporated entity. Second, when Ona-powered persistent execution ships to production Codex users, and which plan tiers get access. Third, how Cursor and GitHub Copilot respond: both have roadmaps that include persistent agent execution, and an OpenAI-owned cloud runtime changes the competitive math.

Sources