How Lover Snap works
Last updated:
How Lover Snap works
1. Industry-leading image engine
Lover Snap is built on Flux, the photorealistic image-generation model by Black Forest Labs. Flux is the same family of models used by professional studios for high-fidelity AI imagery, which is why our outputs look like photographs rather than illustrations or filter-driven selfies.
2. AI character modeling
When you upload reference photos:
- The Flux engine analyzes facial features, body proportions, hair, and identifying characteristics.
- It builds a private AI character scoped to your account.
- Training takes about 30 minutes. You can close the app while it runs; we’ll send a notification when the character is ready.
3. Intelligent scene integration
When you generate a photo:
- You choose a scene (or write a custom prompt).
- The system generates an internal prompt that combines your character with the scene description.
- Flux renders the image. Photos take about 23 seconds each; you can run up to 16 simultaneous generations.
4. Data security and privacy
- Reference photos and trained characters are encrypted at rest (AES-256) and in transit (TLS 1.3).
- Characters are scoped to your account. Nobody else can access them — including us in normal operation.
- We never use your photos or characters to train shared models.
- Deletion: anytime from inside the app. Permanently removed within 30 days.
See Trust and Privacy for the full details.
5. Limits
- Photos of minors are not supported (refused at training time).
- Photos of public figures who aren’t you or a family member are not supported.
- Simulated speech, deepfake video, and audio generation are not supported.
See the AI Content Policy for the full list.