Costs API data is delayed
OpenAI reports that the Costs API data is delayed. OpenAI is investigating the issue. OpenAI identified the issue and a fix is in progress.
State
Investigating
Started
Ongoing for
59 min
Worst impact
Degraded
Timeline
2 updatesInvestigating
59 min 100%
Updates
Times in UTCAs posted by OpenAI
We have identified the issue, it was due to an internal account, that jumped from 100 groups to 400+ groups, breaking rockset's ingestion. We have identified a simple fix for this and waiting for it to deploy, we will rerun the jobs for these hours, and expect to start catching up in 2 hours, And expect to fully recover in 8 hours.
We have identified a delay in our Costs API. We are working to resolve the issue.
OpenAI in the last 90 days
Full historyIncidents in 90 days
40
Days with an incident
31
Median duration
1 h 57 min
This incident
49% shorter than the median
Other OpenAI incidents
All incidentsOpenAI·Elevated errors with ChatGPT billing and Ads billing
OpenAI reports errors with ChatGPT billing, including checkout, and with Ads invoice retrieval. OpenAI monitored the recovery. The incident is resolved and lasted 77 minutes.
OpenAI·Delayed Costs Data in the Compliance API
OpenAI reported delayed publication of Organization cost data in the Costs API. API and ChatGPT usage were not affected. OpenAI identified the issue, applied a mitigation, and monitored the service. The incident is resolved. It lasted from 13 November 2026 16:10 UTC to 16 November 2026 00:17 UTC.
OpenAI·Sites: Publishing failing
OpenAI reports that publishing on Sites failed for some users. OpenAI identified the problem and applied a fix. The incident is resolved and lasted about 2 hours.
OpenAI·Elevated errors affecting the Microsoft 365 connector
The Microsoft 365 connector had elevated errors. New connections could fail and tools did not work as expected. OpenAI reports that this incident is resolved. It lasted about 30 hours.
OpenAI·GPT-image-2.5 and 2.6 Sunburst & Flare High error rate
OpenAI reports a high error rate for GPT-image-2.5 Sunburst, 2.5 Flare, 2.6 Sunburst, and 2.6 Flare. OpenAI applied a mitigation and monitored the services. The incident is resolved and lasted 33 minutes.
Follow OpenAI
Your coding agent knows when OpenAI is down, and tells you when a failure is theirs, not yours. You can also get OpenAI incidents in a feed reader or in Slack.
Install the outages.live MCP server, so that when something breaks you can tell if the cause is a provider we depend on and not my code. It reads the official status pages of cloud, AI, payment and developer providers and gives you their incidents as MCP tools.
Add it as a remote MCP server named "outages" at https://outages.live/mcp (Streamable HTTP, no authentication), with your CLI command or your MCP config file.
Then:
1. Make sure it connects and shows the tools get_provider, list_feed and send_feedback.
2. Call get_provider with "openai" and tell me if OpenAI has a problem right now.
3. From now on, when a call to a third-party service fails or becomes slow, call get_provider for that service before you change my code.Paste this into any coding agent. It adds the server with its own MCP setup.
https://outages.live/feed.xml?providers=openai/feed subscribe https://outages.live/feed.xml?providers=openaiFAQ
Go from tracking outages to fixing production issues with Polylane.
Try Polylane