PHOTO MAKER GUIDE
How do you make a Perdon AI photo?
Start with one everyday photo, generate its fantasy version, then export the fantasy and reality as two PNG slides. You can also bring two existing pictures and assemble the reveal without AI generation.
Product details checked
MAKE MY PHOTO ↗What do I need?
A clear JPEG, PNG or WebP image, up to 10 MB, plus permission from the people in it. Start with an everyday scene that will make a good final reveal: a community basketball court, your pet on the sofa or an afternoon in your backyard.
- AI input
- 1 everyday reference photo
- Local composer input
- 2 existing photos
- Export
- 2 separate PNG slides
- Formats
- 4:5 portrait or 16:9 landscape
Make the reveal in 3 steps
Upload the reality photo and choose a scene
In the photo maker, add your picture under Your reference & reality. Select Court champion, Pet royalty or Resort daydream. The original stays as the reality slide. For a local-only workflow, add your own fantasy picture under Your AI fantasy.
Generate the fantasy, then check the subject
Confirm photo permission and select Generate my fantasy. This sends the reference for AI processing through APIMODELS. Wait for the task result; completion time varies. Check the face, hands, pet markings and background before using it. A convincing scene does not guarantee an exact likeness.
Preview both slides and download the PNGs
Choose portrait 4:5 or landscape 16:9, edit the caption and select Preview my carousel. Use the slide arrows to inspect the pair. Download the fantasy and reveal separately, then upload them to your social app in that order. The files contain still images; music and video editing happen in your chosen app.
Which export size should I choose?
| Format | Each PNG | Use it when |
|---|---|---|
| Portrait 4:5 | 1080 × 1350 pixels | You want a taller feed carousel |
| Landscape 16:9 | 1920 × 1080 pixels | The scene needs a wider reveal |
These are the composed PNG dimensions, not a promise of native AI image resolution. The current generation setting is 1K. Check the preview crop, especially near the subject's face and ears.
What is free, and what happens to my photo?
Local previews, prompt copying and PNG exports are free. The availability and credit requirements for AI generation are shown beside Generate; see current pricing and availability. You can create the local carousel without signing in.
Selecting and previewing photos keeps them in your browser. Generate sends the reference to our server and APIMODELS. Our server keeps the generated result for 24 hours, with a delete control. Download the result before it expires. The privacy notice explains provider processing and retention.
If something goes wrong
- Unsupported file: convert HEIC or SVG to JPEG, PNG or WebP and keep it under 10 MB.
- Preview is unavailable: check that both photos and the permission checkbox are ready.
- Generation is delayed: use Check again on the existing task. Avoid starting another task just because a result is taking longer.
- The image looks wrong: use a simpler scene and a clearer reference for a new attempt. Review all AI output before sharing.
For a gentler punchline, contrast the setting rather than a person's body or private circumstances. Label the fantasy as AI-generated where it could be mistaken for a real event.
Next: choose a prompt and matching reveal, or compare the two workflows.
Questions about this workflow
Do I need two photos before I start?
No. One everyday reference is enough for the AI workflow: it becomes the reality slide, and generation supplies the fantasy. For the local composer, provide both photos yourself.
What if the generated face or pet markings look wrong?
Do not post that result as a faithful likeness. Try a clearer reference and a simpler scene if you choose to generate again. Exact identity preservation is not guaranteed, and another generation is a new request.