OpenAI Outage History

OpenAI experienced elevated error rates affecting image generation functionality across ChatGPT and APIs for 1.1 hours. Users encountered increased errors when attempting to generate or process images. The incident was identified and a mitigation was implemented to resolve the service disruption.

OpenAI experienced elevated error rates that affected ChatGPT users specifically in Europe for approximately 10 minutes. The company quickly identified and applied a mitigation to resolve the issue. The incident was classified as minor and the service was restored with ongoing monitoring of the recovery.

OpenAI experienced an issue where write actions for some ChatGPT workspace connectors were automatically disabled, affecting users' ability to perform write operations through these connectors. The incident lasted approximately 139 hours and was classified as a minor disruption. While OpenAI worked on restoring the affected write actions, workspace administrators could manually re-enable write permissions through the connector settings page as a temporary workaround.

OpenAI experienced elevated error rates specifically affecting ChatGPT Go (5.3 Thinking) users, while other ChatGPT Go models remained unaffected. The incident impacted multiple components including ChatGPT, APIs, file uploads, voice mode, GPTs, image generation, and other services. The issue was resolved after 1.4 hours when OpenAI applied a mitigation and began monitoring the recovery.

OpenAI experienced elevated error rates affecting the gpt-4o-mini model through their API, impacting APIs, Chat, and Responses components. The incident lasted 29 minutes, during which users encountered increased failures when using the affected services. OpenAI identified the issue, implemented a mitigation, and monitored the recovery.

OpenAI experienced intermittent disconnections in the Codex stream, causing elevated errors for users of ChatGPT and Codex services. The incident lasted 31 minutes, during which OpenAI identified the issue, implemented a mitigation, and monitored the recovery.

OpenAI's Sora API experienced elevated error rates affecting some users for 2.3 hours. The service team investigated the issue and applied a mitigation to resolve the problem. The incident was classified as minor and the team monitored the recovery after implementing the fix.

OpenAI experienced elevated errors with the gpt-5.5 model in codex for 53 minutes on April 24, 2026. The incident was classified as minor and affected the specified model's functionality. The issue was under investigation during the reported timeframe.

OpenAI experienced elevated errors affecting Codex gpt-5.5 usage for 2.1 hours, with some users encountering intermittent "not found" errors when starting or updating requests. The incident impacted both Codex and ChatGPT services. The issue was classified as minor and OpenAI worked to mitigate the problem.

OpenAI's Moderation API experienced elevated error rates and latency for 1.3 hours, causing API requests to fail or time out. The issue affected the Moderations API component and was resolved by adding additional capacity to the system. Recovery was gradual, with some requests continuing to experience issues during the mitigation process.