You ask a question, but instead of an answer, you get an "ERROR" message. Here are some the error messages that we are aware of, and what you should do if you see them.
Error occurred (getChatCompletion): Rate limit reached for default-gpt-4 in organization org- on tokens per min. Limit: 40000 / min. Please try again in 1ms. Contact us through our help center at help.openai.com if you continue to have issues.
This means that our system is sending too many requests to the OpenAI model. This is very rare, but has happened before. The only thing you can do is wait a minute or two and try again.
Error occurred (getChatCompletion): This model's maximum context length is 8192 tokens. However, you requested 9792 tokens (1792 in the messages, 8000 in the completion). Please reduce the length of the messages or completion.
You should never see a message like this as our system is fine-tuned to not exceed it's token limits. However, accidents do happen. If you see a message like this, contact us immediately. In the interim, you can usually fix something like this by reducing your "Limit" (Advanced Query Screen) which is the maximum number of context documents the system will return for on your question.
Error occurred (getChatCompletion): Incorrect API key provided: sk-8n85g***************************************87Jn. You can find your API key at https://platform.openai.com/account/api-keys.
Another error you should never see, but if you do see this error, please contact us immediately. OpenAI has a nasty habit of deleting API keys without notice and for no apparent reason.
Error occurred (): That model is currently overloaded with other requests. You can retry your request, or contact us through our help center at help.openai.com if the error persists. (Please include the request ID 225dd54ee337f5b2251d0a1dee0da5fe in your message.)
This means that the OpenAI servers are overloaded. Nothing anyone can do except wait for a few minutes, then try again. This is a error message you could see during peak periods of the day.
We recorded several of these in June 2023, but as of this writing (July 31, 2023) none since then.