Hello OpenAI team, I have noticed that if max_prompt_tokens is set low like ~500 some very strange hallucinations start happening with code interpreter and chat responses. probably due to messages being dropped due to tokens taken up by tool definitons. If the max_prompt_tokens is set to ~10.000 o…

Thanks for highlighting this! We’ll add some documentation around reasonable numbers to use for max_prompt_tokens when using tools.

Possible Assistants API v2 bug with max_prompt_tokens + tools

jknt June 11, 2024, 10:35am 6

This message has more details

Topic		Replies	Views
Assistant API v2: max_prompt_tokens gets exceeded, barely, consistently Bugs	5	840	July 4, 2024
Assistant API run status incomplete with max_prompt_tokens when max_prompt_tokens is NOT set API assistants-api	3	859	June 26, 2024
All my runs are "Status: incomplete" Bugs api , assistants-api	16	1822	November 7, 2024
Too many input tokens are used by Assistant Feedback assistants-api	2	90	November 20, 2024
Unexpected token counts of 800 however i have just created a run API api	3	69	December 13, 2024