4 ms·
As someone using OpenAI in production, I can attest to the lack of stability and consistent performance in the current offering. Depending on the time of the da
by mritchie712 4y ago
As someone using OpenAI in production, I can attest to the lack of stability and consistent performance in the current offering. Depending on the time of the day (and who knows what else), the same calls to GPT3 can take from 500ms to 15 seconds.
- senthilnayagam 4y agochatgpt launched about 2 months back and have over 100 million users, they have secured funding from microsoft so they are going to optimise and scale its infrastructure. go for their paid offering if latency is bothering you
- alvah 4y agoParent is clearly making API calls to GPT3, which isn't really comparable to using ChatGPT (paid or unpaid) via a browser.
- mritchie712 4y agoCorrect.