Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Its not just for coding, the base “gpt-4” model seems better than the latest preview model

https://platform.openai.com/docs/models/continuous-model-upg...



The -turbo models in the past have been much worse too. gpt-3.5-turbo is way way worse than text-davinci-003 (gpt-3.5).

The -turbos are correspondingly priced. gpt-4-turbo is ~1/3 price of gpt-4, 6.6x more expensive than gpt-3.5-turbo-instruct and 20x gpt-3.5-turbo.


I wish base gpt-4 was available in the chat product, miss it.


You can get the baseline GPT-4 model without new system prompts via ChatGPT Classic: https://chat.openai.com/g/g-YyyyMT9XH-chatgpt-classic

It is an official GPT provided by ChatGPT with GPT-4 as backend.


I'm not using ChatGPT now, but isn't this the old GPT-4? https://chat.openai.com/g/g-YyyyMT9XH-chatgpt-classic


Thanks!




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: