OpenAI Outage History

The OpenAI Responses API experienced elevated error rates starting at 11:06 PM UTC and was resolved by 11:56 PM UTC, approximately 50 minutes later. The incident affected the APIs / Responses component, potentially causing request failures or slower responses for some users. OpenAI issued two status updates during the incident, first noting elevated errors and that mitigation was being implemented, then confirming the service had returned to normal. The issue appeared to be limited to the Responses API, with no broader impact reported.

OpenAI's ChatGPT service experienced degraded performance related to retrieving workspace members, impacting 13 components including ChatGPT / Voice mode, ChatGPT / GPTs, ChatGPT / Conversations, ChatGPT / Login, ChatGPT / Compliance API, ChatGPT / Search, ChatGPT / File uploads, ChatGPT / Image Generation, ChatGPT / Deep Research, ChatGPT / Agent, ChatGPT / Codex, ChatGPT Atlas, and ChatGPT / Connectors. Users reported difficulties loading or accessing workspace member data and related features. The issue began at 7:37 PM UTC and lasted for about 3 hours, being resolved by 10:36 PM UTC. The provider published two status updates during the incident as teams investigated and remediated the issue. Some user reports originated from the United States, suggesting localized impact rather than a global outage.

OpenAI's APIs / Responses experienced degraded performance due to a GPT-5.1 Codex Max Elevated Error Rate. Users began reporting issues around 8:30 AM UTC, and the OpenAI status page listed a degraded service for the APIs / Responses at 9:06 AM UTC. The degradation affected the API layer, with higher error rates and slower responses for GPT-5.1 Codex endpoints. The incident lasted approximately 50 minutes, resolving by 9:56 AM UTC. The provider issued 2 status updates during the event and confirmed mitigation was applied as recovery proceeded.

The incident began at Dec 25, 2025 3:57 AM UTC when OpenAI flagged problems loading conversation history in ChatGPT Conversations. Over the next several hours, some users also experienced issues loading history and downloading files from Custom GPTs in ChatGPT GPTs. The outage lasted about 5 hours and 44 minutes, with the affected components returning to normal by Dec 25, 2025 9:41 AM UTC. Users across regions reported degraded access, with past chats not loading and downloads failing during the incident. The two components impacted were ChatGPT / Conversations and ChatGPT / GPTs, reflecting a partial outage with degraded functionality for retrieving past chats and managing GPT assets.

An issue affected a subset of Android users trying to access ChatGPT, resulting in degraded performance across multiple ChatGPT components. The incident began at 8:36 PM UTC and was resolved by 9:37 PM UTC, lasting roughly one hour. The outage impacted 13 components including ChatGPT Agent, Atlas, Conversations, GPTs, Image Generation, Deep Research, Connectors, Search, File uploads, Voice mode, and Login. User reports were scattered, reflecting intermittent access across features.

OpenAI's Sora API experienced elevated error rates starting at 6:16 AM UTC, causing degraded image generation for some users. The incident affected the APIs / Sora component, with reports of service issues beginning in the early UTC window. The outage lasted about 45 minutes, with the service returning to normal by 7:01 AM UTC.

OpenAI Sora API experienced a brief incident with degraded latency and increased error rates affecting the APIs / Sora component. The outage began at Dec 19, 2025 1:21 AM UTC and was resolved by Dec 19, 2025 1:26 AM UTC, lasting roughly 5 minutes. Users attempting to call the Sora API may have observed slower responses or intermittent failures.

OpenAI's status page issued 2 updates during the incident, noting that mitigation had been applied and recovery was being monitored. The Sora API has returned to normal, and the issue appears isolated to this API with no broader OpenAI services indicated.

On 10:15 PM UTC, users reported intermittent loading after signing in to ChatGPT via SSO, indicating a degraded login experience. The issue affected the ChatGPT / Login component and appeared to be global given multiple reports. OpenAI acknowledged the degraded condition and began mitigation, with the status updated to ChatGPT / Login degraded at Dec 18, 2025 11:21 PM UTC.

The incident involved two official status updates and the service returned to normal by 11:26 PM UTC. The short degradation lasted about 5 minutes and impacted users trying to access ChatGPT after SSO authentication.

OpenAI experienced a login outage affecting Sora Web and ChatGPT login via SSO. Users were unable to sign in to Sora Web, and ChatGPT sign-in experiences were intermittently degraded after authentication. The outage began at 9:55 PM UTC and was resolved by 10:51 PM UTC, lasting roughly 56 minutes. The issue appeared to be global, with user reports from around the world. StatusGator detected the issue at Dec 18, 2025 9:55 PM UTC, about 11 minutes before OpenAI logged the official acknowledgment at Dec 18, 2025 10:06 PM UTC.

An authentication disruption affected OpenAI services, preventing sign-in for ChatGPT users via SSO and causing Sora Web login failures. Users across multiple regions reported sign-in errors and stalled or failed login attempts. The outage began at Dec 17, 2025 9:46 PM UTC and lasted approximately 25 hours, impacting two main components: ChatGPT / Login and Sora / Login, with intermittent issues also observed in Finetuning APIs during the window.

OpenAI issued several status updates and implemented mitigations as the incident progressed. Finetuning APIs briefly showed intermittent failures but were restored by Dec 17, 2025 10:11 PM UTC. The Sora Web login problem escalated to an outage and was mitigated with recovery completed by Dec 18, 2025 10:51 PM UTC. By that time, ChatGPT sign-in returned to normal and overall services stabilized. The event drew a global wave of user reports indicating widespread sign-in problems across regions.