OpenAI Outage History

OpenAI's Sora API experienced a brief degradation in latency and increased error rates, impacting clients making requests to the Sora endpoint. The incident began at Dec 19, 2025 1:21 AM UTC and the degraded performance persisted until Dec 19, 2025 1:26 AM UTC when services returned to normal. A small number of user reports indicated issues earlier, around 12:30 AM UTC and 12:45 AM UTC. OpenAI's status page described the event as degraded latency and increased error rates for the Sora API, with mitigation applied and monitoring ongoing. Two status updates were issued during the incident, reflecting the recovery progress.

OpenAI ChatGPT login issues briefly degraded the login experience after SSO authentication. Users began reporting intermittent login loading problems around 10:15 PM UTC. The OpenAI status page shows the ChatGPT / Login component moving from up to warn at Dec 18, 2025 11:21 PM UTC and returning to up at Dec 18, 2025 11:26 PM UTC after mitigation.

OpenAI issued two status updates during the incident, noting that mitigation had been applied and recovery was being monitored. The outage lasted about 5 minutes and primarily affected the ChatGPT login flow, with services restored to normal after the final update.

An authentication outage blocked sign-in to OpenAI's Sora Web, preventing access for users trying to login and use Sora features. The issue affected Sora / Login and later ChatGPT / Login, with sign-in errors reported across regions. The incident began at 9:55 PM UTC and OpenAI published updates including a down state for Sora / Login at 10:06 PM UTC. Mitigation steps were applied and monitoring continued, with the status moving to a degraded state at 10:36 PM UTC and recovery by 10:51 PM UTC.

OpenAI experienced a login outage affecting Sora Web sign-in and ChatGPT login flows. Users reported sign-in errors and authentication failures when attempting to access Sora Web or ChatGPT via SSO, with some receiving errors or blank screens. The disruption impacted two components: Sora / Login and ChatGPT / Login, and appeared to be global as reports came from many regions. The outage lasted about 45 minutes, beginning at 10:06 PM UTC and ending at 10:51 PM UTC.

OpenAI issued multiple status updates during the incident (six updates were observed). The messages described mitigation efforts and monitoring as the team worked to restore sign-in. Updates progressed from initial outage notices to confirmation of mitigations, culminating in restored access for users by the end of the window.

OpenAI reported elevated errors affecting two components: ChatGPT / Codex and APIs / Responses, causing degraded functionality for users attempting Codex workflows or API calls. The incident began at Dec 17, 2025 1:06 AM UTC when the OpenAI status page showed a degraded state and noted that mitigation had been applied and recovery was being monitored. The outage lasted about 3 hours and 10 minutes and was resolved by Dec 17, 2025 4:16 AM UTC.

OpenAI issued two status updates during the incident as they worked to restore normal service, and communications indicated mitigation steps and ongoing recovery. Some users reported issues during the window, reflecting a localized impact on those interacting with Codex features and API responses.

OpenAI reported degraded performance beginning at Dec 16, 2025 6:46 PM UTC as Codex CLI began experiencing elevated errors. A follow-up update at Dec 16, 2025 10:36 PM UTC expanded the issue to the Codex & Responses API, with impacts to ChatGPT / Codex and the API/Responses. Users attempting code generation, chat interactions, or API calls encountered errors or degraded responsiveness, with reports of heavy load and missing responses from multiple regions. The disruption lasted approximately 5 hours and 55 minutes, and recovery began by Dec 17, 2025 12:41 AM UTC as affected components returned toward normal.

OpenAI issued three status updates during the incident, describing mitigation steps and ongoing monitoring. The issue affected two main components: ChatGPT / Codex and APIs / Responses, indicating a broad impact on both interactive chat/code generation and programmatic access. User reports came from diverse regions around the world, including North America, Europe, and the Asia-Pacific, suggesting wide geographic reach. By the end of the window, the status page indicated recovery and continued monitoring while services stabilized.

OpenAI's Image API experienced a brief degradation, with image-related endpoints returning errors as the Images component showed a high error rate. The incident began at Dec 16, 2025 4:16 AM UTC when the Images API moved to a degraded state, and engineers applied mitigation as they monitored recovery. The disruption lasted about 5 minutes and was resolved by Dec 16, 2025 4:21 AM UTC. The issue was confined to the APIs / Images component and did not indicate a broader service impact.

The outage began on Dec 9, 2025 10:51 PM UTC and was resolved by Dec 12, 2025 11:17 PM UTC, affecting OpenAI's ChatGPT connectors across global users and causing degraded performance and errors across many features. The incident impacted at least 12 components, including Conversations, Login, Compliance API, Search, Codex, Deep Research, Agent, File uploads, Voice mode, GPTs, Image Generation, and ChatGPT Atlas. Users reported failures, timeouts, and incomplete responses, with Android users also seeing elevated errors. The outage lasted about 3 days and 26 minutes before stabilization.

OpenAI experienced a brief degradation affecting Codex Web and Code Reviews within ChatGPT / Codex. Users attempting to use these features reported intermittent issues, with reports coming from Europe (including the United Kingdom and the Netherlands). The incident lasted about 55 minutes. The issue began at 8:26 PM UTC and was resolved by 9:21 PM UTC. OpenAI issued two status updates during the incident to communicate the investigation and restoration of services.

OpenAI reported elevated errors impacting ChatGPT Voice mode, with users experiencing login issues and error messages. The issue began to be observed before the official acknowledgment, and the OpenAI status page transitioned the ChatGPT / Voice mode from up to a degraded state at Dec 9, 2025 7:36 PM UTC before returning to normal at Dec 9, 2025 8:11 PM UTC. The outage lasted about 35 minutes, and affected users attempting to access voice-related features.