OpenAI buys Pierre, the small team behind Codex’s file tree and diff viewer

OpenAI said on November 4 that it will acquire Pierre Computer Company, the maker of Code Storage and the open-source Trees and Diffs libraries. According to the official announcement, the goal is to extend Codex with better tools for building, understanding and reviewing code.

The announcement gives no price, no deal value and no closing date. It says the deal is subject to customary closing conditions, including regulatory approvals, and that until it closes OpenAI and Pierre remain separate and independent companies.

Libraries Codex already runs on

OpenAI says anyone who has browsed files or reviewed a change in the Codex app has already used Pierre’s work. Trees handles navigating a project’s files, and Diffs shows what changed. OpenAI describes both as built for reviewing large volumes of code with high performance.

The announcement lists other Pierre projects too. Diffs has an Edit mode that lets a reviewer fix a line right where the problem appears. Pierre also recently released Highlights, a syntax highlighter that, per OpenAI, keeps up while a developer types or an agent streams code. The third piece is Code Storage, which gives other applications a way to keep their code and its full change history in Git repositories.

For current users, OpenAI makes two commitments. Pierre’s open-source projects will stay open source, and OpenAI says it will keep contributing to them. Customers can keep using Code Storage after the deal closes, while OpenAI works to bring its features into Codex. The announcement does not say how long Code Storage will stay available as a standalone product, or whether its pricing will change.

What Code Storage was built to do

The official post is thin on Code Storage, so Pierre’s own launch post fills in the picture. Published on October 14, 2025, it describes a white-label, low-latency Git layer that developers embed in their own applications, aimed at teams building AI and code-generation products. According to Pierre, repositories can be created on demand without rate limits, and the service offers AI-oriented features such as ephemeral branches and cold storage.

Pierre also made performance claims in that post. It said the system can support millions of new repositories per day and that reads and writes are about 60 times faster than comparable solutions built on S3 or R2 object storage. These are the company’s own figures, and neither Pierre’s post nor OpenAI’s announcement offers independent verification.

Source control for agents becomes a product

The deal makes more sense next to OpenAI’s engineering post from earlier this week, which OpenAI cites. It describes Pando, an in-house source control backend for engineers and agents. OpenAI says it has seen how the systems around Git need to change as more work happens in parallel. We could not read that post directly, so the details of Pando are not covered here.

The pieces fit together in a specific way. Pando is an internal system, Code Storage is a service other companies already build on, and Trees and Diffs are the review interface. Together they cover storing code, tracking changes and reviewing them. OpenAI states its aim as to “set the standard for modern software development”, from the first branch an agent creates to deployment. That puts OpenAI in territory held by code hosts and review tools, which until now have been built around human developers.

It also fits a pattern in AI coding. In April, SpaceX’s $60 billion bet on Cursor showed how much value is being placed on the tools where coding agents work. OpenAI’s purchase is far smaller in ambition on its face, but it targets the plumbing underneath those tools. It is also another small-team deal after OpenAI’s purchase of Orchestra Bio last month.

What the announcement leaves open

Several questions remain. OpenAI has not said how many people are joining, what it is paying, or when the deal might close. It does not say whether Code Storage will keep its current form or fold into Codex, only that its features and functionality will move there. The stated ambition to “advance the future of source control across the industry” is not backed by any product details or timeline. For now, Codex users see no change, Code Storage customers can carry on, and the open-source libraries keep their licenses and public repositories.