OpenAI is experiencing a partial outage that has affected access to ChatGPT, Sora, and the OpenAI API, according to the company’s official status page.
When Did It Start?
The issue began late Monday night, and by 5:30 AM PT Tuesday, OpenAI had identified the cause and began working on a solution. However, by 8:00 AM PT, the company warned that full recovery may take several more hours.
This means that many users—especially on the U.S. West Coast—may not be able to access ChatGPT this morning.
Impact on Users
- Elevated errors and slow response times are expected
- Some users received a “Too many concurrent requests” error when trying to use GPT-4o
Unlike typical outages, which are often resolved quickly, this one has been notably longer in duration.
What Else Is Happening?
The outage comes at a busy time for OpenAI:
- On Monday, Apple announced new integrations with OpenAI’s models during its WWDC event.
- Also on Monday, OpenAI confirmed it had reached $10 billion in annual recurring revenue.
- On Tuesday morning, CEO Sam Altman announced an 80% price cut for developers using the new o3 AI models via API.
Scaling Challenges
OpenAI has faced ongoing challenges in scaling its infrastructure to support its rapidly growing user base. CEO Sam Altman previously joked that the company's "GPUs are melting" under the demand, highlighting the strain on computing resources.
