OpenAI doubles Pro limits and pools ChatGPT and Codex usage into one allowance

What OpenAI announced

OpenAI has confirmed three changes to how its products handle usage, according to BleepingComputer. The company is doubling the usage limits on its Pro plan. ChatGPT and Codex will draw on a single combined usage allowance. And the Work capabilities that previously lived in a separate experience are being folded into the regular Chat interface.

A dev.to write-up covering the report notes that the announcement is thin on detail, which leaves several questions open.

One allowance for two products

The merged ChatGPT and Codex allowance is the change most likely to alter daily workflows. Many developers use ChatGPT for planning, research and debugging while handing implementation work to Codex. If both now consume from one pool, a long research session in ChatGPT could shrink what remains available for a Codex coding run, and the reverse.

That shared-pool reading is an inference drawn from the reporting, not a confirmed detail. BleepingComputer does not explain how the combined allowance is measured, whether it counts messages, tokens, compute credits or something else, so the mechanics of how one product's usage affects the other's headroom remain unknown. Checking how consumption appears in account settings once the change lands will be the fastest way to learn the rules.

Work features arrive in Chat

Moving Work capabilities into regular Chat appears designed to collapse a separate mode into the main conversation interface, so users no longer jump between distinct experiences. The reporting does not describe what those features include, so anyone considering them should test on low-stakes work before folding them into an established routine.

What the reports leave unanswered

Neither the original report nor the follow-up coverage specifies:

• the concrete figures behind the old and new Pro limits, making the practical size of the increase impossible to judge
• whether plans other than Pro are affected
• how the combined allowance is calculated
• a timeline for when every user sees the changes

dev.to's coverage explicitly flags these gaps and points to OpenAI's own channels for exact numbers.

Why it matters

Usage caps shape developer workflows in a direct way. Hitting a limit mid-task forces a pause at exactly the wrong moment, and for teams those caps feed into tooling budgets. Doubling Pro limits helps heavy users in principle, but without knowing the starting figures the practical impact stays an open question.

The pooled allowance cuts both ways. It could add flexibility, letting unused ChatGPT headroom cover Codex work and the reverse. It could also create a new failure mode, where research in one product crowds out capacity needed in the other. Which effect dominates depends on how OpenAI calculates and surfaces combined usage, a detail nobody has published yet.

The Work-in-Chat change is lower stakes but points in the same direction: OpenAI appears to be consolidating its product surfaces rather than maintaining parallel experiences.

What to do meanwhile

Pro subscribers should track their consumption in the weeks after the rollout to gauge the difference. Anyone relying on both ChatGPT and Codex should confirm how the pooled allowance is counted before planning work around it. Teams budgeting for AI tooling should revisit their assumptions rather than treat the changes as cosmetic. Until OpenAI publishes specifics, the exact limits, affected plans and timing all remain unresolved.