How photo-to-macros actually works

A few people asked how the "snap a photo, get macros" flow works in Fud AI.

It's less magic than it looks:

1. You take the photo (or pull one from your library).
2. The image goes to a vision model with a structured prompt.
3. It comes back as estimated calories + macros, which you can edit before saving.

Key word is estimate. It's a starting point you nudge, not a scale. Still way faster than typing "grilled chicken, ~140g, plus rice, plus...".

More on the BYOK side tomorrow. App: https://www.fud-ai.app/