OpenAI committed a change to use a different http method which doesn’t directly support timeout (but can close connections). Setting a parameter directly on the library still works though.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Simple Request exeeds Completion Token Limit of 4096 Token; GPT-35-turbo | 3 | 1171 | November 22, 2023 | |
| ChatGPT answers partially to request | 6 | 365 | February 20, 2025 | |
| Is it me or GPT4 consistently doesn't finish and cuts the answers? | 18 | 6921 | April 11, 2024 | |
| Text translation with finish_reason length | 2 | 2806 | February 23, 2024 | |
| Gpt-4o-transcribe truncates the transcript | 15 | 2589 | August 29, 2025 |