OpenAI Outage History

OpenAI experienced elevated latency and errors across their APIs, Chat, Embeddings, and Audio services for 2.5 hours. The company investigated the issue, implemented a mitigation solution, and monitored the recovery. All affected services were restored to normal operation.

OpenAI experienced a 6-minute outage where users were unable to access chat support through the help center, with Login and ChatGPT components affected. Users encountered errors when attempting to reach chat support and were directed to email support instead. The issue was identified and a mitigation was implemented to resolve the problem.

OpenAI experienced login issues specifically affecting FedRAMP users who had recently logged out, impacting access to ChatGPT and login services. The incident lasted 23.3 hours, during which users were unable to log back into their accounts after logging out. OpenAI identified the root cause, implemented a mitigation, and monitored the recovery to restore normal login functionality.

OpenAI experienced an issue where an increased number of users were hitting Codex rate limits, affecting both Codex and ChatGPT services. The incident lasted 18.4 hours before a mitigation was successfully applied. The team monitored the recovery after implementing the fix.

OpenAI experienced elevated latency and error rates affecting ChatGPT 5.5 Thinking functionality on both web and ChatGPT platforms. The incident lasted 1.4 hours and was resolved by applying a mitigation, with the team continuing to monitor the recovery.

OpenAI's ChatGPT paid plans experienced elevated error rates for 1.6 hours, affecting both web and ChatGPT services. Users reported being logged out of their accounts and unable to reconnect during the incident. The issue was resolved after OpenAI implemented a mitigation and monitored the recovery.

OpenAI users experienced increased error rates when using GPT-5.4 and GPT-5.5 models through the API and Chat services. The incident affected API responses and lasted approximately 1.1 hours before a mitigation was successfully applied and the service recovered.

OpenAI experienced performance degradation issues with GPT 5.5 that affected some users for 32.1 hours. The service team investigated reports of worse performance and applied a mitigation to resolve the issue. The incident was classified as minor and the team monitored the recovery after implementing the fix.

OpenAI's Codex Cloud and Code Review services experienced elevated error rates for 32 minutes. Users encountered increased failures when attempting to use these code-related services. The issue was identified and a mitigation was implemented to resolve the problem.

OpenAI experienced a major service incident affecting ChatGPT access for logged-out users and login functionality. The issue prevented users from accessing ChatGPT without being logged in, impacting the service's availability for anonymous users. The incident was resolved after 26 minutes when OpenAI applied a mitigation and confirmed recovery.