On November 4, 2026, OpenAI announced an agreement to acquire Pierre Computer Company, Inc., with the aim of bringing Pierre’s Code Storage service into Codex. OpenAI says Codex already uses two of Pierre’s tools, Trees and Diffs. The announcement said the deal’s closing was subject to customary conditions, including regulatory approvals, and that the companies would remain independent until closing. OpenAI’s announcement
What Pierre’s tools do in Codex
OpenAI says Codex already uses Trees to help developers navigate project files and Diffs to view code changes. Diffs also has an Edit mode that lets a reviewer fix a line in place.
OpenAI describes Pierre’s Highlights as a syntax-highlighting tool that updates as a developer types or an agent streams code. The company also describes Code Storage as a service that lets apps store code and track its history using Git. Its plan is to bring Code Storage’s features and functionality into Codex after the acquisition closes.
What OpenAI says about customers and open source
OpenAI says Code Storage customers can continue using the service after closing while integration work proceeds. It also says Pierre’s open-source projects will remain open source and that OpenAI will continue contributing to them. These commitments cover both users of the hosted service and developers working with the open-source projects.
How Code Storage and Pando fit different roles
Code Storage is described as a service for apps; Pando is OpenAI’s internal, Git-compatible backend for development in its monorepository. The descriptions place them in different roles: Code Storage handles app-facing code storage and history, while Pando supports OpenAI’s own repository.
In a November 2 engineering post, OpenAI reported that Pando had nearly fivefold lower median push latency and delivered nearly one petabyte of Git data per day. Those figures describe Pando, not Code Storage. OpenAI’s engineering post on Pando
Bin He Talks offers English-language commentary on the deal and its Git infrastructure, including the host’s view that code review is becoming a bottleneck as AI agents produce code.
