OpenAI Outage History

OpenAI reported a brief outage affecting multiple API endpoints, including /v1/chat/completions, /v1/completions, Private Link, and the /Responses API. Users attempting to use chat completions or text completions may have encountered errors or timeouts during the incident. The issue degraded two components: APIs / Responses and APIs / Completions (legacy).

The OpenAI finetuning API outage blocked the creation of new finetune jobs via the API. Customers relying on custom model training were unable to start new finetune runs, disrupting workflows that depend on model customization. The incident began at Jul 8, 2025 12:16 AM UTC and the service remained degraded from Jul 8, 2025 1:22 AM UTC until full recovery at Jul 8, 2025 6:56 PM UTC.

OpenAI issued 3 status updates during the incident, documenting the outage, the mitigation applied, and the recovery progress. Mitigation was put in place and monitoring continued as OpenAI worked to restore normal operation of the APIs / Fine-tuning component. A small volume of user reports described authentication-related issues during the outage, reflecting the broader disruption while the service was degraded.

The OpenAI ChatGPT / Codex component experienced elevated error rates, causing degraded performance for Codex-powered features within ChatGPT. Users attempting Codex-powered tasks may have encountered errors or timeouts during the incident. The outage began on 9:36 PM UTC and was resolved by 10:17 PM UTC, lasting about 41 minutes.

The provider issued 2 status updates during the incident, describing mitigation steps and the recovery. The service returned to normal after the final update.

OpenAI reported a brief degradation of API endpoints beginning at Jul 7, 2025 6:01 PM UTC with elevated error rates across 15 API components, including Realtime, Responses, Login, Fine-tuning, Assistants, Audio, Batch, Embeddings, Files, Images, Completions (legacy), Chat, Compliance, Moderations, and Uploads. Users could experience failed requests or timeouts across multiple OpenAI services during the window. The incident affected a broad portion of the API surface, indicating widespread impact across the platform.

The issue lasted about 36 minutes and OpenAI posted 2 status updates during the incident. All components returned to normal by Jul 7, 2025 6:37 PM UTC. The event involved a broad set of API components, but overall user-reported activity remained low during the window.

On Jul 2, 2025 6:37 PM UTC OpenAI flagged an outage relating to Pay-As-You-Go API accounts experiencing 429 errors tied to negative balance checks. The issue impacted 15 API components, including Chat, Completions (legacy), Embeddings, Login, Files, Images, Moderations, and other core APIs, leading to degraded access for affected users. The incident lasted about 3 hours and 44 minutes, and OpenAI published two status updates during the event. Services were restored by 10:21 PM UTC.

OpenAI Sora briefly went offline, affecting Image Generation, Video Generation, Video Viewing, and Feed. Users encountered elevated errors during the incident, with service degraded for about 24 minutes. The outage began at Jul 1, 2025 4:47 PM UTC and was resolved by Jul 1, 2025 5:11 PM UTC.

OpenAI issued 2 status updates as the team investigated and then restored functionality.

The outage began on Jun 23, 2025 10:06 PM UTC with Compliance API data delays. By Jun 24, 2025 8:16 PM UTC the issue had expanded to a broad set of OpenAI APIs, causing elevated errors and degraded responsiveness across multiple services including Realtime, Responses, Embeddings, Fine-tuning, Login, Images, Chat, and more. A subset of Sora services also experienced outages during the period. The incident generated a large volume of user reports and affected more than a dozen components.

OpenAI issued multiple status updates over the course of the incident as mitigation steps were implemented. The situation gradually improved, with most services returning to normal by Jun 29, 2025 12:16 AM UTC. The outage lasted approximately 5 days and 2 hours.

OpenAI ChatGPT / Conversations experienced a brief degradation that affected paid users, with elevated errors during conversations. The outage began at Jun 19, 2025 2:11 AM UTC and was resolved by Jun 19, 2025 2:17 AM UTC, lasting about 6 minutes. Users may have seen errors or intermittent responses during the chat interactions.

OpenAI issued two status updates during the incident. The first noted high error rates for paid users and that a mitigation was being implemented. The second confirmed that the issue had been resolved and that normal operation had resumed for ChatGPT / Conversations.

OpenAI reported a degraded login experience when some users saw a 'Content failed to load' message after signing in, affecting the ChatGPT / Login component. The outage began on Jun 18, 2025 8:06 PM UTC and was resolved by Jun 19, 2025 12:21 AM UTC, lasting about 4 hours and 15 minutes. Affected users could not load content after login, limiting access to chat features in affected sessions.

OpenAI released two status updates during the incident, first noting an investigation into the login-related content loading issue and later confirming restoration of normal login content loading. A reported example from a user in San Diego described chatGPT not responding during the outage.

OpenAI reported a brief degraded period affecting ChatGPT / Conversations, with users experiencing elevated errors. The incident began at Jun 18, 2025 1:21 AM UTC and was resolved by Jun 18, 2025 1:37 AM UTC, lasting about 16 minutes. The OpenAI status page posted two updates during the incident, describing conversation errors and then confirming restoration. User reports showed a noticeable spike in issues around the middle of the event, with impact focused on ChatGPT / Conversations.