Sam Altman resets OpenAI Codex limits
OpenAI CEO Sam Altman announced that the company is resetting usage limits for Codex, its cloud-based software-engineering agent, to mark crossing 3 million weekly users. In a post on X, Altman wrote: "To celebrate 3 million weekly codex users, we are resetting usage limits. We will do this every million users up to 10 million. Happy building!" A reset wipes the portion of a quota a developer has already consumed in the current period, restoring full access for the remainder. Business Today reports the milestone is up from about 2 million users roughly a month earlier, underscoring how quickly agentic coding tools are scaling. The repeated-reset pledge doubles as a growth incentive and a way to keep heavy users engaged while OpenAI manages the underlying compute and capacity demands that Codex usage places on its infrastructure.
What happened
OpenAI CEO Sam Altman said the company is resetting usage limits for Codex, its cloud-based software-engineering agent, after the product crossed 3 million weekly users. Announcing the move on X, Altman wrote: "To celebrate 3 million weekly codex users, we are resetting usage limits. We will do this every million users up to 10 million. Happy building!" Business Today reports that the reset clears the share of a quota a developer has already used in the current window, so even someone who had nearly exhausted their allowance starts again from zero for the rest of the period.
Why it matters
The 3 million figure is up from roughly 2 million weekly users about a month earlier, per Business Today, a steep ramp that illustrates how fast agentic coding assistants are being adopted. Tying a perk to user milestones is a familiar growth tactic, but for an agent that executes multi-step tasks against real repositories it is also a demand-management lever: resets encourage continued engagement from the most active developers without OpenAI permanently raising published limits.
The practitioner read
For teams that depend on Codex in their workflow, the load-bearing detail is operational, not promotional. Usage caps and resets shape how reliably a coding agent can be embedded in CI, batch refactors, or long-running tasks. Cloud coding agents are compute-intensive, and provider economics, not model quality alone, increasingly govern how much capacity reaches users. The pattern to watch across this category is whether generous-access promotions are followed by tighter steady-state limits, tiered pricing, or queueing once a milestone push ends.
What to watch
- •Whether OpenAI sustains the every-million-users reset cadence toward the stated 10 million target, or quietly slows it.
- •Any move to formalize higher limits, new pricing tiers, or rate controls for Codex once growth stabilizes.
- •Reliability and latency signals during usage spikes, which indicate whether capacity is keeping pace with adoption.
Key Points
- 1Altman reset Codex usage limits to celebrate 3 million weekly users, pledging another reset at every additional million up to 10 million.
- 2Each reset restores a developer's consumed quota for the current period, effectively widening access during a fast user ramp.
- 3Heavier free or near-free agent usage raises OpenAI's serving and capacity costs, a tension every coding-agent provider now manages.
Scoring Rationale
A usage-limit reset tied to a 3-million-weekly-user milestone is a meaningful access and growth signal for developers who rely on Codex, but it is an incremental product-operations change rather than a frontier model or platform shift. It sits in the solid-but-not-major band.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
