Back to home

Closed on Sunday AI Photo Requirements: Selfies, Full-Body Photos and Pets

What photo to upload: formats, size, lighting, occlusion, multi-person photos and pets — and which results are verified vs just advice.

Last updated: 2026-10-22

Short answer: one clear photo — JPG, PNG or WebP up to 10MB — with the subject facing the camera, well lit, and nothing covering the face. One subject per photo works best; pets should be full body.

The hard requirements

These are enforced or strongly checked by the generator:

  • Formats: JPG, PNG or WebP (the file must really be that format — renamed files are rejected).
  • Size: up to 10MB per photo.
  • Consent: the person in the photo must be you or an adult who agreed. No minors, no public figures, no sexual or harassing content — you confirm this before every generation.
  • One subject: the generator replaces the central subject. If several people fill the frame, expect unpredictable results — crop or pick a single-subject shot.

Selfies and portraits (people)

Verified to work well:

  • Facing the camera, face clearly visible
  • Good, even lighting — golden-hour photos are fine, dark club photos are not
  • Head to waist in frame, so the outfit carries into the video
  • Arms free (the trend pose crosses them anyway)

Usually not worth the credit:

  • Sunglasses, hats or hair covering half the face
  • Mirror selfies — the phone can appear in your video
  • Heavy filters or beauty apps that alter the face
  • Extreme close-ups (face only, no shoulders) or far-away full-body dots

Pet photos

  • Full body in frame works best; the pet becomes the central subject (choose the Pet in the choir template) or joins a cat choir around a person
  • Pet facing the camera, eyes visible
  • On the ground or a simple surface — a hand-held pet gets a hand in the video
  • Same formats and size limits as people photos

See the pet tutorial for the two pet modes explained with examples.

Verified results vs advice

  • Verified: a standard front-facing phone selfie of one person produced the results on the examples page — that's the baseline you can expect from a decent photo.
  • Advice: the pet and low-light guidance above reflects how image-to-video models generally behave; we label it as advice rather than a guarantee, because AI results vary with every photo.

Ready? Upload your photo — the generator checks format and size before you can spend anything.