ChatGPT Outage Hits Globally
- ChatGPT suffered a widespread outage across multiple regions while OpenAI announced a large funding round. - The incident disrupted developer and production workflows that depend on hosted LLM services. - The outage highlights operational dependence on hosted models and the need for reliable fallbacks and local tooling (tradingview.com).
OpenAI said on Monday, April 20, that some users were unable to load ChatGPT and Codex as a global outage spread across its consumer and developer tools. (status.openai.com) The company posted the incident at 7:35 a.m. Pacific time and classified it as “degraded performance,” with 12 ChatGPT components and one Codex component affected. (status.openai.com) User reports surged at about 10:05 a.m. Eastern time, according to Invezz, which cited Downdetector data showing peaks above 8,000 reports in the United Kingdom and roughly 1,700 to 1,875 in the United States. (invezz.com) OpenAI said users could not load ChatGPT and Codex, while outside reports said the failures hit login, failed responses, and interrupted writing and debugging on web and mobile. (status.openai.com) (moneycontrol.com) ChatGPT is OpenAI’s consumer chatbot, and Codex is its coding product for turning prompts into software tasks. When both are down at once, the outage reaches office chat windows, software teams, and customer-facing tools that sit on top of OpenAI’s hosted systems. (openai.com) (9to5mac.com) The disruption landed less than three weeks after OpenAI said it had closed a $122 billion funding round at an $852 billion post-money valuation on March 31. OpenAI said the round was anchored by Amazon, Nvidia, and SoftBank, with continued participation from Microsoft. (openai.com) In that same announcement, OpenAI said ChatGPT was nearing 1 billion weekly active users and that Codex was changing how developers turn ideas into working software. Those claims put extra weight on any outage that blocks access during the workday. (openai.com) OpenAI’s status page says ChatGPT logged 99.85% uptime from January through April 2026, while Codex showed 100% uptime over the same period on the public dashboard. The page also says those figures are aggregate metrics and that individual customer availability can vary by tier, model, and feature. (status.openai.com) The company had not published a root cause or a repair timeline in its initial incident updates on Monday. By late morning, the story was still the same as the lede: a service many people now treat like infrastructure was not loading. (status.openai.com)
