Image generator

User prompts → DALL-E or Replicate → stored gallery. ~30 minutes.

A user types a prompt, picks a model, generates an image, sees it stored in their personal gallery.

Time: ~30 min. Integrations: OpenAI (DALL-E) or Replicate.

Full walkthrough being written.

Foundation prompt

An image generator. User types a prompt; picks model (DALL-E 3 or Replicate flux);
clicks Generate. Image saved to their gallery. Gallery view shows thumbnails;
click for full size; download button.

Use OpenAI's DALL-E 3 by default; allow switching to Replicate's flux models.
Store generated images in Convex file storage.

Related

OpenAI

DALL-E.

Replicate

Open-source models.

File storage

Storing the results.