OpenAI Outage History

OpenAI experienced an incident affecting ChatGPT Login, ChatGPT Web, and the APIs / Responses, causing degraded access and API performance for some users. Users reported difficulty signing in, incomplete page loads, and slower or failed API responses. The disruption impacted three components: ChatGPT / Login, ChatGPT / Web, and APIs / Responses.

OpenAI Responses API and Files API experienced a brief degradation. Users may have encountered slower responses or intermittent errors when calling these endpoints. The OpenAI status page listed degraded performance for the two endpoints and said the team was investigating, with two status updates issued. The outage began on Apr 30, 2025 1:41 AM UTC and was resolved by Apr 30, 2025 2:04 AM UTC, lasting about 23 minutes. The impact was limited to these two components, with services returned to normal afterwards.

The platform.openai.com outage began at Apr 24, 2025 7:22 PM UTC and was resolved by Apr 24, 2025 7:57 PM UTC, affecting the Playgound and related tooling. In the days that followed, OpenAI users experienced elevated error rates and degraded performance across ChatGPT features, APIs, and related services, including image generation, moderation, DALL·E, and video capabilities. The impact was global, with users reporting login failures, slow responses, and content-generation issues across multiple regions. The incident unfolded over several days with regular status updates and mitigations being applied by OpenAI.

OpenAI experienced a brief disruption affecting retrieval of stored completions and access to GET /v1/chat/completions, which prevented viewing stored completions in the OpenAI dashboard. The issue was isolated to the Chat API component and caused elevated errors during the incident. The outage began at Apr 8, 2025 7:11 AM UTC and was resolved by Apr 8, 2025 11:16 AM UTC.

The incident affected a broad set of OpenAI ChatGPT components on 2025-04-07. The outage began at 3:01 PM UTC, with 16 ChatGPT components listed as degraded at the peak, including ChatGPT / 1-800-CHAT-GPT, Context connectors, Dall-E, File uploads, GPT vision, Voice mode, Memory, WhatsApp, Search, Advanced Data Analysis, Single Sign-On, Image Generation, Login, Web, and Mobile apps for iOS and Android. Users reported impaired access and slower responses, with at least one report from Kuala Lumpur around 3:30 PM UTC underscoring the global reach of the disruption.

OpenAI's status page reported a brief degradation affecting the ChatGPT / GPT Vision component from Apr 5, 2025 8:46 AM UTC to Apr 5, 2025 9:56 AM UTC, lasting about one hour. The incident was acknowledged with two official updates before service returned to full functionality, and the impact appeared limited to the ChatGPT / GPT Vision features during that window.

OpenAI Labs experienced a temporary outage causing experimental features and tools in the Labs environment to be unavailable. The outage began on 1:51 PM UTC and was resolved by 3:41 PM UTC, lasting about 1 hour and 50 minutes. OpenAI's status page shows the Labs component transitioning from up to down and back to up during the window. Two official status updates were published during the incident, describing the impact and recovery progress. User reports indicated issues during the outage.

On Apr 3, 2025 1:11 AM UTC, OpenAI Sora Image Generation and Sora Video Generation transitioned to a degraded state. The two components returned to normal at Apr 3, 2025 1:21 AM UTC. User reports appeared earlier, with observations at 12:45 AM UTC and 1:00 AM UTC, suggesting users experienced issues before the official acknowledgment. The incident lasted about 10 minutes and affected generation tasks in both Sora components.

OpenAI issued two status updates during the event, documenting the degradation and restoration. By Apr 3, 2025 1:21 AM UTC, the components were listed as up and services returned to normal. The impact was limited to the Sora image and video generation features with no broader service impact indicated.

OpenAI experienced a degradation impacting multiple components across ChatGPT and Sora. Users reported upstream connect errors and failures to generate responses, with some interactions timing out. Affected components included ChatGPT / Web, ChatGPT / Mobile - iOS, ChatGPT / Mobile - Android, and Sora services such as Login, Image Generation, Video Generation, Video Viewing, and Feed, totaling eight components. The issue appeared globally, with user reports from Salt Lake City, London, Barcelona, Mumbai, and beyond indicating widespread impact. The outage began on 2:12 PM UTC and continued for several hours.

OpenAI reported a brief degraded performance for ChatGPT across Web and mobile clients. The issue began at 8:37 AM UTC and was resolved by 9:14 AM UTC, spanning roughly 37 minutes. Users attempting to access ChatGPT on the Web, iOS, or Android could have experienced slower responses or difficulty loading conversations during this window.