OpenAI Outage History

The outage began on Apr 24, 2025 7:22 PM UTC and affected multiple OpenAI components, including platform.openai.com (Playground) and a broad set of ChatGPT services and APIs. Users experienced degraded performance and intermittent outages across ChatGPT web and mobile interfaces, image and video generation features, Moderation and DALL·E APIs, and related tools such as Conversations, Memory, and Single Sign-On. The incident spanned several days with ongoing updates and mitigations as OpenAI worked to stabilize the platform and recover functionality.

Customer reports during the period indicated widespread login failures, slow or failing chat responses, and errors when generating images or videos. Some users also encountered upstream connection errors and general service instability across global regions. The OpenAI status page documented elevated error rates and progressive mitigation attempts across multiple services as the company coordinated recovery efforts.

OpenAI issued numerous status updates throughout the incident, describing investigations into elevated errors, implementation of mitigations, and continuous monitoring. By late April, key services began to recover; by Apr 29, 2025 10:57 PM UTC, components such as the Video generation service (Sora) returned to normal operation and core platform access gradually stabilized.

OpenAI experienced a partial outage impacting the APIs / Chat when users attempted to view stored completions in the dashboard or execute the GET /v1/chat/completions endpoint. The issue began at 7:11 AM UTC and was resolved by 11:16 AM UTC, lasting roughly 4 hours and 5 minutes.

OpenAI experienced a brief degradation affecting multiple ChatGPT components starting at 3:01 PM UTC. A total of 15 components including core chat features, DALL-E image generation, memory, file uploads, web access, search, and mobile apps were listed as degraded, causing slower responses and intermittent issues for users. Recovery began for many components around 3:41 PM UTC, and by 6:36 PM UTC all listed components were back to normal. The outage lasted roughly 3 hours and 35 minutes.

The provider issued 3 official status updates during the incident, documenting the transition from degraded to restored status across affected components.

OpenAI reported a brief degraded performance affecting ChatGPT and GPT Vision. The incident began at 8:46 AM UTC and was resolved by 9:56 AM UTC, lasting about 1 hour and 10 minutes. Users interacting with ChatGPT or GPT Vision may have experienced slower responses or intermittent issues during the degraded period.

OpenAI published two status updates during the incident to describe the degraded state and progress toward remediation. The issue appeared isolated to the ChatGPT and GPT Vision components, with no broader impact reported on other OpenAI services during this window.

OpenAI Labs experienced a brief outage affecting the Labs component, causing some users to encounter failures or degraded access to lab-related features. The outage began on Apr 4, 2025 1:51 PM UTC and was resolved by Apr 4, 2025 3:41 PM UTC, lasting about 1 hour and 50 minutes. OpenAI posted two status updates during the incident. The impact appeared limited to the Labs component, with other OpenAI services remaining operational. A user reported issues from Coimbatore, India.

OpenAI's Sora / Image Generation and Sora / Video generation components experienced a brief degraded performance. Users attempting to generate images or videos during the incident likely faced slower processing or intermittent errors. The issue began at 1:11 AM UTC and was resolved by 1:21 AM UTC, for a total duration of about 10 minutes.

OpenAI's status page showed two updates during the incident, first marking the components as degraded and then confirming restoration to normal. While there were multiple user reports, the impact appeared limited to that short window and did not indicate a broader outage across other OpenAI services.

The incident impacted eight OpenAI components, including ChatGPT Web, ChatGPT mobile apps (iOS and Android), and Sora services (Login, Image Generation, Video Generation, Video Viewing, and Feed). Users reported errors and slower responses as the degraded performance persisted across these services.

OpenAI reported a short period of degraded performance for ChatGPT across Web and mobile clients, beginning at 8:37 AM UTC and resolving by 9:14 AM UTC, roughly 37 minutes. The incident affected ChatGPT Web, iOS, and Android access, and the status page issued two official updates during the event before full restoration was confirmed. No additional details about the cause were provided in the public updates.

OpenAI ChatGPT experienced degraded performance across 16 components, impacting a broad range of features including messaging integrations (ChatGPT / WhatsApp, ChatGPT / 1-800-CHAT-GPT), vision and image generation (ChatGPT / GPT vision, ChatGPT / Dall-E, ChatGPT / Image Generation), memory, data analysis, search, and web access. Users likely faced intermittent access or reduced responsiveness across these features. The degradation lasted about 1 hour and 4 minutes, from Apr 2, 2025 1:07 AM UTC to Apr 2, 2025 2:11 AM UTC.

The outage began at Apr 1, 2025 9:11 PM UTC and was resolved by Apr 1, 2025 10:01 PM UTC. The degradation affected two components, ChatGPT / Login and ChatGPT / Web, likely causing login difficulties and slower access to the ChatGPT web interface during the window. The incident lasted about 50 minutes before services returned to normal.