Here you find OpenAI system status and ChatGPT errors:
OpenAI Status - Incident History Statuspage
- Elevated errors on DALL·E APIon September 29, 2023 at 4:43 am
Sep 28, 21:43 PDTResolved - This incident has been resolved.Sep 28, 18:45 PDTMonitoring - A fix has been implemented and we are monitoring the results.Sep 28, 18:32 PDTInvestigating - We are investigating an issue affecting some customers of the DALL·E API.
- High error rates on DALLE API and Labs generationson September 26, 2023 at 10:49 pm
Sep 26, 15:49 PDTResolved - This incident has been resolved.Sep 26, 15:42 PDTMonitoring - A fix has been implemented and we are monitoring the results.Sep 26, 15:39 PDTUpdate - The fix has been implemented and we are continuing to monitor the situation.Sep 26, 15:21 PDTIdentified - The issue has been identified and a fix is being implemented.Sep 26, 15:10 PDTInvestigating - We are currently investigating this issue.
- Degraded performance on the Embeddings APIon September 26, 2023 at 12:57 am
Sep 25, 17:57 PDTResolved - Between 8am - 4pm PT, a misconfiguration of our internal systems resulted in degraded performance of roughly 20% of embeddings requests. The error introduced may have impacted downstream applications relying on vector similarity. We recommend recomputing any embeddings produced between 8am - 4pm PT.Sep 25, 15:05 PDTInvestigating - As of 8am PT, the Embeddings API is returning degraded outputs. We are currently investigating.
- Elevated error rates for sampling requests for fine-tunes of davinci-002 and babbage-002on September 21, 2023 at 12:27 am
Sep 20, 17:27 PDTResolved - This incident has been resolved.Sep 20, 16:35 PDTMonitoring - A fix has been implemented and error rates are lowering to normal background levels.Sep 20, 15:43 PDTIdentified - The issue has been identified and a fix is being implemented.
- Moderation API returning incorrect scoreson September 20, 2023 at 6:00 pm
Sep 20, 11:00 PDTResolved - Moderation results are back to pre-incident correctness.Sep 20, 10:56 PDTMonitoring - We've updated the moderation service to a previous working version and are monitoring the rollout.Sep 19, 03:30 PDTInvestigating - As of September 19th around 3pm, the Moderation API is returning incorrect scores. We are currently investigating.