We are pleased to announce that we have simplified our copywriting and article writing use case guidelines, including a significant relaxation of the token and rate limits. The key things to know are:
- We have increased the maximum user input tokens and output tokens to 250 (previous input and output token limits varied by feature but were generally around 50 input tokens and 20-150 output tokens)
- We have increased the end user rate limit to 60 requests per minute (previous rate limits varies by feature but were generally between 5 -18 generations per minute)
- We require the use of the content filter and end user IDs
These changes come as a result of our frequent review of risks as well as improvements in our safety mitigations. Please stay tuned for further updates to the other sections of our use case guidelines!
As always we are grateful for your support in our commitment to safe and responsible AI. We hope these updates will come as a welcome change and we can’t wait to see what you build!
Please reach out if you have any questions