OpenAI Outage History

The OpenAI Responses API experienced a brief degradation in its Web Search tools, with elevated error rates affecting Web Search functionality within the Responses API. The incident began on Sep 5, 2025 9:37 AM UTC and was resolved by 9:51 AM UTC, lasting about 14 minutes. OpenAI documented the incident on its status page, noting a degradation in the APIs / Responses component and that a mitigation was being implemented. The provider issued 2 status updates during the incident to communicate the degradation and recovery. The issue appeared localized to the Web Search tools within the Responses API and did not indicate broader service impact.

OpenAI experienced a brief degradation in the File API surface (APIs / Files) beginning at Sep 4, 2025 9:36 PM UTC after elevated errors were reported for file-related operations. The issue affected the APIs / Files component and caused errors for file uploads, downloads, and other file-related API calls. The degraded service lasted about 45 minutes and was resolved by Sep 4, 2025 10:21 PM UTC.

OpenAI communicated the incident via its status page, issuing two updates and confirming that mitigation steps were underway. The status returned to normal by Sep 4, 2025 10:21 PM UTC.

There was a small volume of user reports indicating issues, suggesting the disruption was not broad in scope.

OpenAI experienced a widespread degradation in which ChatGPT failed to display responses across multiple interfaces. Users reported blank replies, prompts not returning, and missing responses across ChatGPT web, mobile apps, and related features. The incident affected 23 components, including File uploads, Search, Dall-E, Memory, Conversations, Login, Web, and various mobile clients. The issue began with user reports around 7:15 AM UTC and was acknowledged by OpenAI at Sep 3, 2025 8:11 AM UTC with a note that investigators were looking into the problem.

OpenAI ChatGPT Conversations briefly experienced elevated errors impacting user interactions with the ChatGPT Conversations component. The incident began at 3:27 PM UTC and was resolved by 3:46 PM UTC, lasting about 19 minutes. OpenAI published two status updates during the incident, first noting an investigation into the elevated errors and later confirming that the issue had been resolved and normal service had resumed.

OpenAI reported elevated errors affecting ChatGPT / Conversations. The incident began at 12:26 AM UTC and was resolved by 1:01 AM UTC, lasting about 35 minutes. Engineers implemented mitigations as they investigated the cause. The ChatGPT / Conversations component was the focus of the incident, and the issue impacted user interactions during the window.

OpenAI issued two status updates during the incident, describing the mitigation efforts and the eventual restoration of normal operation.

OpenAI experienced a period of degraded login affecting ChatGPT and API access beginning at Aug 28, 2025 2:16 PM UTC. Users reported login failures and timeouts, with two components impacted: ChatGPT / Login and APIs / Login. The outage lasted approximately 1 hour and 10 minutes and was resolved by Aug 28, 2025 3:26 PM UTC.

OpenAI issued three official status updates between Aug 28, 2025 2:16 PM UTC and Aug 28, 2025 3:26 PM UTC detailing the mitigation progress and eventual restoration of login services. Reports came from users across multiple regions, indicating broad impact rather than a localized issue.

The OpenAI ChatGPT Search feature experienced issues affecting a subset of users, with search requests failing and elevated errors reported. The outage began at Aug 27, 2025 8:02 PM UTC and lasted until Aug 28, 2025 12:51 AM UTC, for an overall duration of approximately 4 hours and 49 minutes. Free-tier users were noticeably affected, while other ChatGPT functionalities remained available.

OpenAI's status page documented the incident with two official updates. The ChatGPT / Search component was listed as degraded during the event and later returned to normal by Aug 28, 2025 12:51 AM UTC.

OpenAI reported degraded performance across multiple API endpoints, including APIs / Responses, APIs / Completions (legacy), and APIs / Chat, with users experiencing elevated errors and latency. The outage began on 10:41 PM UTC and was resolved by 1:56 AM UTC, lasting approximately 3 hours and 15 minutes. The incident affected three API components and prompted three official status updates as OpenAI worked to mitigate the issue.

OpenAI's status updates detailed the progression: the first update at Aug 26, 2025 10:41 PM UTC announced elevated errors and latency and that mitigation was being implemented; by Aug 26, 2025 11:06 PM UTC some endpoints had recovered (Completions (legacy) and Chat) while others remained degraded; the final update at Aug 27, 2025 1:56 AM UTC indicated the last affected component (APIs / Responses) had returned to normal.

OpenAI's ChatGPT Voice mode experienced degraded performance, with users reporting that voice conversations dropped previous messages and elevated errors across the affected services. The issue impacted the ChatGPT / Voice mode component and disrupted voice-enabled chats for a period.

OpenAI experienced a brief period of degraded API performance across multiple services, with elevated errors and latencies observed on the APIs / Embeddings, APIs / Audio, and APIs / Completions (legacy). The outage began at Aug 22, 2025 6:36 PM UTC and was resolved by Aug 22, 2025 6:46 PM UTC, lasting about 10 minutes. OpenAI issued 2 status updates during the incident and worked on implementing a mitigation as they investigated the cause. During the window, users may have encountered failed requests or slower responses when interacting with the affected APIs, after which services returned to normal.