Brand Models
Train MomentA on your own images so generated artwork matches your style or subject.
Most tools forget who you are between projects. Brand Models learn from a set of your images so new artwork is generated with your look — or your signature subject — baked in.
Two kinds of model
Style
Teach a look — color, texture, lighting, and art direction — so every new image matches your aesthetic.
Subject
Teach a recurring character, product, or object so it stays consistent across scenes, poses, and lighting.
Train a model
- 1
Choose the intent
Decide whether you're teaching a Style or a Subject.
- 2
Upload images
Add 8–20 images (4 minimum). For style, vary subjects but keep the look consistent; for subject, vary angles and backgrounds.
- 3
Train
MomentA processes your images and builds the model. This runs in the background.
- 4
Use it anywhere
Once ready, apply the model in Create so artwork is generated in your style or with your subject.
Image guidance
Use 8–20 clear, high-quality images. More consistent inputs produce a stronger, more reliable model.
Note
Custom brand (LoRA) models are a Pro feature and above. Generating images requires the AI image capability, which is available on adult accounts and can be granted to children by a guardian.