# What differentiates "redacted" ChatGPT from raw Davinci?

**URL:** <https://community.openai.com/t/what-differentiates-redacted-chatgpt-from-raw-davinci/219113>\
**Category:** ChatGPT\
**Tags:** chatgpt\
**Created:** [May 19, 2023, 6:12pm UTC](https://community.openai.com/t/what-differentiates-redacted-chatgpt-from-raw-davinci/219113 "2023-05-19T18:12:50Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![shredengineer](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/shredengineer/32/133501_2.png) [@shredengineer](https://community.openai.com/u/shredengineer)\
**Post date:** [May 19, 2023, 6:12pm UTC](https://community.openai.com/t/what-differentiates-redacted-chatgpt-from-raw-davinci/219113/1 "2023-05-19T18:12:50Z")

</div>

Obviously, ChatGPT has several filters and fine-tunes in place. This includes filtering profanity and inserting disclaimers. On the other hand, requests to the vanilla Davinci model give much more raw answers.

I want to know the exact details of the different operating conditions between both models.

Is there any info publically available?