OpenAI's Codex Hits 4 Million Weekly Developers, Launches Enterprise Push

OpenAI says Codex now has more than 4 million developers using it every week, up from more than 3 million in early April — a gain of roughly a million users in just two weeks.

The company disclosed the figures in an announcement focused on the next phase of the product: enterprise adoption at scale. OpenAI is launching Codex Labs, a program that embeds OpenAI experts inside organizations through hands-on workshops and working sessions, and it is partnering with seven global systems integrators — Accenture, Capgemini, CGI, Cognizant, Infosys, PwC, and Tata Consultancy Services — to accelerate deployments.

The stakes are straightforward. AI coding assistants have become the fastest-moving category in enterprise software, and weekly developer counts are the primary competitive metric among vendors. But turning individual usage into production deployments inside large organizations is a harder problem, and that is the gap OpenAI says its new programs address.

From individual tool to enterprise workflow

OpenAI reports that enterprises are deploying Codex across the software development lifecycle, with a roster of named customers. Virgin Atlantic uses Codex to increase test coverage and team velocity, which the company says reduces technical debt and improves performance. Ramp uses it to accelerate code review. Notion uses it to build new features quickly. Cisco applies it to understanding and reasoning across large, interconnected repositories. Rakuten uses it for tasks such as incident response.

The pattern OpenAI describes follows a familiar adoption curve: "What starts with one team often expands quickly as leaders see the gains in speed, output, and leverage."

Codex is also expanding beyond writing code. OpenAI says the product now supports browser-based work, image generation, memory, and ongoing work across tools and apps. Teams use it to pull context from different systems, reason through what matters, and produce outputs such as briefs, plans, checklists, drafts, and follow-ups — then act on them. That broadens the addressable market well beyond engineering teams.

"That creates a larger opportunity for enterprises to help every team move faster — not just the teams writing code," OpenAI said.

Codex Labs: embedding experts inside enterprises

The new Codex Labs program brings OpenAI specialists directly into customer organizations. According to the announcement, enterprise teams learn where Codex fits, how to integrate it into existing workflows, and how to move from early usage to repeatable deployment. OpenAI frames the goal plainly: "help enterprises get real value from Codex, faster."

Organizations can apply through a form on OpenAI's site or contact their OpenAI account team.

A channel strategy built on GSIs

The systems integrator partnerships are the more strategically significant move. OpenAI states the reason directly: "The demand we're seeing is outpacing our ability to help enterprises adopt Codex as quickly as they'd like."

The seven partners — Accenture, Capgemini, CGI, Cognizant, Infosys, PwC, and TCS — bring something OpenAI lacks at scale: experience operating inside large enterprises. As the announcement notes, these firms "know how to modernize software delivery, integrate new systems, support change across complex organizations, and help customers move from pilot to production."

The partners will help enterprise customers identify and deploy high-value Codex use cases across the software development lifecycle. They are also eating their own cooking: each firm is using Codex internally, reworking how their own teams operate to build repeatable delivery models they can bring to customers.

This channel approach mirrors how enterprise software has historically scaled. Direct vendor teams can service a limited number of accounts; a global SI network extends reach into thousands of enterprises that already trust those integrators with transformation budgets. For OpenAI, it converts raw adoption numbers into something more durable: production deployments embedded in how companies build software.

Why the numbers matter

The jump from 3 million to 4 million weekly developers in two weeks signals accelerating adoption at a time when competition in AI coding tools is intense. The metric covers individual developers, and OpenAI did not break down how many are enterprise users versus free or individual subscribers. The company also did not disclose Codex pricing or contract values tied to the new enterprise programs.

What OpenAI does claim is direction: "Codex is already becoming part of how enterprises build software and run production workloads." The company says it wants partners to "help us scale that work globally."

The test ahead is execution. Codex Labs is an application-based program with unspecified capacity, and OpenAI itself admits demand exceeds its ability to support adoption directly. Whether the GSI partnerships can move customers from pilots to production at pace will determine whether the 4 million weekly developers figure becomes a floor for growth or a peak of curiosity.