Vtry AI docs
Learn the complete Vtry AI workflow in the interface, then use the API reference to build the same person, clothes, try-on, and edit flow from your backend.
Interface workflow
The interface follows the same order as the API: add a person, add clothes or accessories, generate a try-on, then edit the result when you need a focused correction.
Step 1
Create the person profile that Vtry AI will preserve across try-ons and edits.
Open the Persons area, click Add Person, and enter a unique name for the model.
Use clear full-body or half-body photos. Front, side, and angled references usually produce the most stable identity.
Step 2
Build a reusable clothes and accessories library from uploads, image URLs, product pages, or search.
Open Clothes and choose the input method that matches your source.
Use Extract when the item is worn by a model and Use as is when the image is already a clean product shot.
Step 3
Compose a full try-on using one person, one or more selected items, and a scene instruction.
In Try-On Studio, select the person, then choose the clothes or accessories that should appear in the final image.
Use a custom prompt, Prompt Gallery preset, or Random prompt. Strong prompts describe where the person is, what they are doing, and how the item should be visible.
Step 4
Refine an existing generation, a gallery image, or a clothes image with a direct edit instruction.
Open the Edit tab and select either a gallery result or a clothes image as the source.
Keep edits simple and explicit, such as change the coat to navy, add a slight smile, or show the person skiing.
Step 5
Share people, clothes, and generation history with teammates on a shared workspace.
Open your profile, go to Team Workspace, create an invite link, and share it with teammates.
Paid plans
Advanced generation model
Use Nano Banana 2 when you need more precise multi-item styling, stronger prompt following, higher resolution output, or better preservation of visible product details.
Layer multiple distinct clothing items and accessories in one generation.
Follow detailed scene, pose, preservation, and styling instructions more reliably.
Produce sharper, richer outputs for polished ecommerce and editorial images.
Preserve visible typography, logo-like details, patterns, and small material cues better.
Build with the same workflow
The API tab maps these same interface steps to endpoint calls, request bodies, polling, and response formats.
