Conjure takes a detailed text prompt and turns it into an image, with the account layer handled entirely by Google OAuth — no separate signup flow. Every generation a user makes gets saved automatically to their personal library with a one-click download, no extra save step required. A follow-up note to the original brief asked for a public “bento” gallery feed on top of that, showing everyone's generations by default, with a per-image toggle so a specific creation can be kept out of the public feed without needing to be deleted afterward. There's no generation cap for now, per the same follow-up note.
The prompt
A webapp that generates an image from a prompt. Prompt can be very detailed, so as to guide the model. User can save image and download it. Simple Google oAuth auth for account creation and login. (Follow-up: added a public “bento” gallery feed, with a per-image private toggle.)
What it does
- Text-to-image generation with support for long, detailed prompts
- Personal gallery of previously generated images per user
- One-click image download
- Google OAuth account creation and login
Built in 8 minutes, prompt to live deploy.