OpenAI Outage History

OpenAI experienced elevated error rates and degraded latency with the Codex GPT-5.5 engine, affecting both ChatGPT and Codex services for 2.1 hours. Free users encountered high error rates while all Codex users experienced slower response times. The issue was mitigated and the service was restored to normal operation.

OpenAI's Realtime API experienced an outage affecting SIP/WebRTC flows, impacting APIs and Audio components. The incident lasted 3 hours and was classified as minor severity. The service was under investigation throughout the duration with no resolution details provided in the available updates.

OpenAI experienced elevated error rates with GPT 5.5 that affected APIs, Chat, and Responses for 3.2 hours. The issue was identified as users experiencing elevated errors across the impacted services. A mitigation was applied and the service was being monitored for recovery.

OpenAI experienced elevated errors affecting ChatGPT file uploads and Codex Cloud task creation for 4.3 hours. The issue impacted users' ability to upload files to ChatGPT and create tasks in Codex Cloud. OpenAI applied a mitigation and monitored the recovery to resolve the problem.

OpenAI experienced degraded performance with Codex Cloud tasks, where follow-up prompts were failing at higher than expected rates. The incident affected both ChatGPT and Codex services for 1.2 hours. The team investigated the issue, though the summary does not specify the final resolution method.

OpenAI experienced increased error rates and higher than normal latency for the gpt-5.5 model API, affecting APIs, Chat, and Responses components for 1.8 hours. The service team implemented mitigations that gradually reduced error rates and restored normal performance. The incident was resolved after applying final mitigations and confirming recovery through monitoring.

OpenAI experienced elevated transcription failures that affected ChatGPT, Voice mode, and Codex services for 15.3 hours. Users encountered increased error rates across these services during the incident. The issue was resolved by applying a mitigation, and the service was returned to normal operation under monitoring.

OpenAI experienced increased error rates with image generation through their API, affecting the Images component and broader API functionality. The incident was classified as minor and lasted approximately 1.8 hours. The issue was under investigation with monitoring in place to track the problem.

OpenAI experienced an issue that made the ChatGPT pricing and overview pages unavailable for 1.7 hours. The company investigated the problem, implemented a mitigation solution, and restored service to the affected website pages.

OpenAI's Responses API experienced elevated error rates affecting API functionality for 1.2 hours. The issue was investigated and a mitigation was applied, with the service being monitored for recovery.