OpenAI Codex Usage Limit Resets: Tracking and Trends

OpenAI Codex Usage Limits are Frequently Reset

OpenAI frequently performs manual resets of usage limits for Codex and ChatGPT Work users, often as compensation for technical failures or to celebrate growth milestones. A community tracker, Codex Resets, monitors these events, recording 35 total resets with an average interval of 8.9 days. These resets restore users to 100% of their weekly and hourly limits, sometimes providing "banked resets" that users can save and apply manually via the web, mobile, or desktop applications.

Primary Drivers for Usage Resets

Based on the announcement logs from OpenAI representative @thsottiaux, resets are typically triggered by four main categories of events:

1. Technical Failures and Outages

Resets are used as a primary remediation tool when users are negatively impacted by system instability. Examples include:

Cache Optimization Errors: A rollback of a cache optimization that impacted cache hit rates during long sessions led to a full reset for all accounts.
Cyber Risk False Positives: An issue causing requests to be blocked as "high cyber risk" was resolved with a subsequent rate limit reset.
General Outages: Global outages and incidents affecting Codex reliability (such as three separate incidents in 24 hours) are typically followed by usage resets to "make up" for the disruption.

2. User Growth Milestones

OpenAI uses resets to celebrate rapid scaling. The tracker highlights a period of extreme growth where active users across Codex and ChatGPT Work grew from 7 million to 9 million in just a few days. Each milestone (7M, 8M, and 9M users) was accompanied by a usage limit reset or the addition of banked resets to user accounts.

3. Feature Launches and Model Updates

Resets are deployed to encourage experimentation with new capabilities:

Model Adoption: Resets have been issued to allow users to explore "the boundaries of GPT-5.6 Sol" and "GPT-5.1-Codex-Max."
Plugin Launches: Limits were reset to allow users to experiment with newly launched plugins.
Promotional Periods: OpenAI previously lifted usage limits to 2X the usual amount through January 1st as a holiday gift.

4. Infrastructure Scaling

When provisioning additional GPUs to handle slowdowns, OpenAI has reset limits to compensate for the reduced performance experienced by users during the scaling process.

User Perspectives and Strategic Implications

Community discussion on Hacker News reveals a divide between users who appreciate the value and those who view the resets as a strategic "trap."

The "Anchoring" Effect

Some users argue that frequent resets create an artificial usage baseline. By removing the 5-hour usage limit and providing constant resets, OpenAI may be anchoring users to a high-volume workflow that would be unsustainable under standard pricing.

"Those resets and the removal of the 5-hour usage limit are quietly anchoring me to a much higher usage baseline... if one day they just stop doing so, my 'normal' workflow will suddenly exceed the limit."

Competitive Positioning

Users have compared OpenAI's generosity to other providers like Anthropic (Claude) and Google. Some suggest that OpenAI is "burning money" to gain mindshare and developer goodwill ahead of a potential IPO, while others see it as a competitive tactic to keep users within the American model ecosystem.

Operational Frustrations

Despite the free compute, some developers expressed frustration over the lack of programmatic access to usage data. There is currently no canonical API endpoint to check remaining quota or reset times, forcing users to rely on manual checks or unofficial workarounds via the Codex App Server.