Hi
I want to build a subscription model where user will be allowed to interact with my gpt (say 1M tokens / characters) per month on a yearly subscription… i will trace how many token every user used and restrict the reposnse if he exceeds that and give him option to top up
How can i acheive it
Hi @mustaqattari and welcome to the community!
Currently there is no fine-grained control over per-user token usage on your published GPT. OpenAI is currently looking into the monetization scheme for GPTs, which is more like a profit-sharing setup (see here).
So if you want to monetize at per-user-per-X-tokens level, you need to look into a different solution. Either setting up your system from ground up and using OpenAI’s standard ChatCompletions API and accounting per-user tokens with tiktoken. Or setting up something with Assistants and adding an intermediate layer that does the token accounting.
Hope that helps!
1 Like
Welcome to our community!
When starting a business using the API, it is essential to ensure adherence to best practices regarding API keys, build your business model in a way that complies with OpenAI’s terms of service, and deploy services accordingly.
The number of tokens consumed by users can be retrieved from the response of API calls.
Additionally, while skills are required for managing subscribed users, selecting platforms for service provision, and imposing service limitations, it is possible to build these systems.
You must decide for yourself what kind of infrastructure to build on, what billing revenue model to adopt, and what value to provide to users.
For your reference, here are some links to the terms and policies:
https://openai.com/policies/business-terms/
https://openai.com/policies/usage-policies/
Please note that there are areas that require caution, such as healthcare and finance, so if you are providing services related to regulated industries, it is recommended to contact OpenAI’s legal department.