OpenAI rolls out pay-as-you-go Codex pricing for ChatGPT Business and Enterprise teams

What OpenAI announced

On April 2, 2026, OpenAI said teams on ChatGPT Business and Enterprise can now add Codex-only seats with pay-as-you-go pricing. The company’s announcement is straightforward: teams no longer need to buy a fixed Codex seat for every user just to run a pilot. Instead, they can grant full Codex access and pay based on token consumption.

That shift matters because it lowers the organizational threshold for adoption. A manager can now put Codex in the hands of a smaller engineering group, prove value in a few workflows, and then expand usage without locking the company into a seat model designed for broad deployment from day one. OpenAI explicitly frames the change that way, describing a path from small pilots to broader rollout.

What changed in the pricing model

OpenAI said Codex-only seats have no rate limits and are billed on usage rather than a fixed monthly fee. The company also said standard ChatGPT Business seats will continue to include Codex usage limits, which preserves the existing bundle for teams that want wider ChatGPT access instead of a Codex-specific rollout.

• OpenAI lowered the annual price of ChatGPT Business from $25 to $20 per seat.
• Eligible Business workspaces can receive $100 in credits for each new Codex-only team member who joins and starts using Codex, up to $500 per team.
• OpenAI highlighted the Codex app for macOS and Windows, plus Plugins and Automations, as part of the broader push to make Codex easier to connect with existing team systems.

Why this is a meaningful signal

The strongest part of the announcement is the adoption data OpenAI attached to it. The company said more than 9 million paying business users already rely on ChatGPT for work, more than 2 million builders use Codex every week, and the number of Codex users inside Business and Enterprise has grown 6x since January. Those figures do not independently prove long-term retention or ROI, but they do show OpenAI sees enough demand to change packaging and pricing around team deployment.

An inference from OpenAI’s own framing is that Codex is moving from an individual developer product toward a budgeted team capability. The new pricing model is less about selling a premium seat and more about making agentic development easier to test, measure, and scale inside organizations. That is usually what happens when a product owner believes demand is no longer the main constraint and procurement friction becomes the bigger one.

There is still an obvious caveat. This is a company-issued product and pricing announcement, not an external adoption study. OpenAI is describing its preferred commercial model. Even so, the update is high-signal because pricing changes of this kind typically follow internal evidence about how customers are already trying to buy and deploy the product. In that sense, the April 2 X post is less a marketing detail than a signal that Codex is being packaged as an operational tool for teams rather than a side experiment for individual users.

Sources: OpenAI Developers X post · OpenAI announcement