Is anyone aware of any issues with the API service yesterday (Feb 1st)
We were running tests and our daily token usage has been an average of 300,000
Yesterday our token usage spiked massively to 8.2M tokens or a 27x increase over the daily average. There’s nothing we were doing in terms of creating embeddings or running queries so we can’t explain that massive spike on our side.
When we raised this issue with OpenAI they credited us back the cash equivalent of the usage spike which we appreciate but found a little suspicious and maybe indicative that they knew there was a production issue yesterday.
Anyone have any thoughts on this?