OpenAI expands Codex to enterprises with new labs program and global systems integrator partners

OpenAI is pushing Codex deeper into enterprise software work, saying on April 21, 2026, that more than 4 million developers now use the tool weekly and that it is launching a new Codex Labs program alongside partnerships with major global systems integrators.

Codex usage tops 4 million weekly developers

The company said Codex had crossed 3 million weekly developers in early April and climbed to more than 4 million two weeks later. OpenAI framed that growth as evidence that the product is moving beyond individual adoption and into organizational deployment, with companies already using it across testing, code review, feature development, repository analysis, and incident response.

OpenAI said Virgin Atlantic is using Codex to increase test coverage, Ramp to accelerate code review, Notion to build features more quickly, Cisco to reason across large repositories, and Rakuten for incident response. The company also said Codex is expanding beyond coding into browser-based work, image generation, memory, and tasks that span tools and apps.

Codex Labs is designed to speed enterprise rollout

To accelerate adoption, OpenAI is launching Codex Labs, a program that brings company experts into organizations for workshops and working sessions. The aim is to help teams identify where Codex fits, connect it to existing workflows, and move from early use to repeatable deployment.

OpenAI said demand is outpacing its ability to support enterprise adoption on its own, which is why it is also working with a group of large systems integrators including Accenture, Capgemini, CGI, Cognizant, Infosys, PwC, and Tata Consultancy Services. Those firms are expected to help customers move Codex from pilots into production-ready deployments.

OpenAI is betting on software engineering as a commercial wedge

The announcement shows where OpenAI thinks the clearest enterprise value is emerging: software delivery. Coding tools are one of the few parts of the AI market where companies can measure productivity gains quickly, integrate deployments into existing engineering stacks, and justify spending with operational results rather than abstract model benchmarks.

That makes Codex strategically important for OpenAI as it looks to broaden revenue beyond chat products and keep pace with rivals that are also targeting developer workflows. The company’s pitch is no longer just that Codex helps write code faster, but that it can become part of how enterprises build software and run production workloads.

From developer assistant to workflow infrastructure

OpenAI said the broader opportunity now extends beyond engineering teams, with Codex being used to pull context from different tools and turn it into briefs, plans, checklists, drafts, and follow-ups. In practical terms, that pushes the product toward a more embedded role in enterprise operations, where the test is not novelty but whether it can reliably reduce cycle time across daily work.

The company said it wants more organizations to move from limited trials to repeatable deployments, a sign that Codex is becoming one of OpenAI’s main commercial channels for agentic software work.