I’m exploring the possibility of having one custom GPT model in the OpenAI team environment invoke another via actions. Is this technically feasible within the current platform capabilities, and if so, how can it be set up, particularly in terms of permissions and custom action configurations? Any insights or examples would be greatly appreciated!
1 Like
Nope. You could theoretically invoke an Assistant through: https://platform.openai.com/docs/assistants/overview
assistants are like custom gpts, just with an API.
While it’s not possible now, I have been wondering if the @-mentions feature might evolve into something GPTs can invoke themselves.
1 Like
I hope this will be possible soon! A demanding task for infrastructure, but it would be a win/win for all: more use cases, more costs.