How to take photos with multiple people in them
Train a separate model per person, generate solo photos of each, then merge them with Combine Photos on the Max plan or higher.
Each AI model is trained on one person, so a direct prompt with several trained people in one image is not the recommended approach. Instead, create each person separately and merge them.
The recommended workflow
- Train an AI model for each person, using 10-30 photos of that person only.
- Generate a strong solo photo of each person in a similar setting and style.
- Use Combine Photos (Max and Ultra) to merge two photos into one scene. Each combine costs 2 credits.
Plans and models
Starter includes 1 AI model, Pro 3, Max 10 and Ultra 50. Combine Photos requires Max or Ultra.
Being realistic
Group images are harder for AI than single portraits. Results are best when the source photos are clear, similarly lit and similarly framed. Expect to generate a few variations and pick the best one.