Creating images
Turn your selfie and a visual reference into a ready-to-post promo image, and understand what each generation costs.
Every image starts from two inputs: your main selfie, and a reference that supplies the look.
Choose your reference
In Create you can either:
- Pick an Inspiration — a curated look from the Inspiration library. Use this when you do not know what to make next.
- Upload your own reference — any image whose composition, pose, mood, or clothing you want to borrow.
You can also write a prompt, or combine a reference with a prompt.
What each input controls
This split matters, and it is the thing most people get wrong at first:
| Input | What it supplies |
|---|---|
| Your selfie | Identity and face — the only source, always |
| Reference | Composition, pose, mood, clothing |
| Prompt | Adjustments and direction |
A reference never changes who is in the image. Your selfie is the only identity and face source.
Generate
Pick an image size and tap Generate. Each successful manual or Autopilot image uses 4 Credits from the same balance.
All new generations use Seedream. Manual and Autopilot generations are not priced differently.
Where images go
Completed generations stay in Gallery under Creations.
Gallery keeps three things together:
- My Selfies — your uploaded selfies
- Creations — your completed images
- Liked Inspiration — the looks you saved
Privacy
Inputs and generated images are stored privately. The application returns only short-lived access to the signed-in owner. StarSmith does not publish or sell your files, and does not use your images to train a personal model.
Next
Add words to a finished image in Captions.