A: We don't have a public API yet — it's something we're actively planning, and you're not the first person to ask for it. I can't give you a timeline, so I won't make one up.

The part we're working through isn't the endpoint itself, it's usage controls — an API means a script can run through a lot of generations quickly, so the quota and rate-limiting model has to be settled first.

If you're open to sharing: what would your assistants be doing with it — generating at volume, pulling results into another system, or something else? That kind of detail genuinely helps us prioritize it.