- VISUAL
- STATUS
- DATA
- HUMAN
Short answer
How can I do virtual clothing tryon with my AI?
You provide your AI with an image of a garment and an image of a person. The AI uses Kling AI to map the clothing onto the model's body. You get a new image showing the item being worn.
The outcomes
What you get from a request
Depending on how you prompt your agent, you'll get one of these results.
VISUAL
Generated tryon images
The AI processes the garment and model images to produce a final visual of the clothing being worn. This is the primary output for fashion visualization.
STATUS
Processing updates
Since image processing takes time, your agent can check on the progress of the task. You'll know exactly when the render is finished.
DATA
Task tracking
Your agent keeps track of the specific job IDs for every tryon request. This allows you to manage multiple garment mappings at once.
HUMAN
Manual adjustments
The AI handles the mapping, but you decide if the fit or lighting looks right. You remain in control of the final creative direction.
The workflow
What your AI does when the request arrives.
Your agent acts as the bridge between your instructions and the Kling AI processing engine.
Initiate mapping
Your agent sends the garment and model images to the engine to start the overlay process.
virtual_try_onTrack progress
The agent monitors the background processing to see when the image is ready.
get_tryon_taskVerify completion
Once the engine finishes the render, the agent confirms the task status is complete.
get_tryon_taskDeliver output
The agent presents the final mapped image or the task details back to you in your chat interface.
get_tryon_task
Try it
Copy these to start.
Use these prompts to trigger the Kling AI tools immediately.
Starting points
Swap the bracketed text with your actual image files or specific garment descriptions.
Kling AI Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_ouNEUpfXe2olXwoUZxOeCTNQmBbguqmUtasfkq7h/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Kling AI capabilities are ready to use.
{
"mcpServers": {
"kling-ai-generative-video-image-mcp": {
"url": "https://edge.vinkius.com/vk_preview_ouNEUpfXe2olXwoUZxOeCTNQmBbguqmUtasfkq7h/mcp"
}
}
}Perform a virtual tryon using [garment_image.jpg] on [model_image.jpg].
Check the status of my last tryon task using the ID [task_id_123].
Map this [blue silk blouse] onto the person in [model_photo.png].
Use the virtual_try_on tool to see how this [denim jacket] looks on [model_v2.jpg].
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Kling AI once, then ask.
Just click the button to link your account. Your credentials stay encrypted, and you can start using the tools in your preferred AI client right away.
Connect Kling AI to your AIFAQ
How this task behaves
- 01
Can the AI change the model's physical features?
No. The tool maps the clothing onto the existing person in the image without altering their identity or body structure.
- 02
Does the AI automatically save the images to my computer?
No. The AI generates the images and provides them to you in the chat. You must manually save any results you want to keep.
- 03
Can I try on multiple items at once?
Yes. You can trigger multiple tryon tasks, and your agent can track each one individually.
- 04
What happens if the clothing doesn't fit the model's pose?
The AI attempts to warp the garment to match the model's shape, but extreme poses may affect the realism of the result.
- 05
Can the AI edit the background of the image?
No. The tool is specifically for mapping garments onto models; it does not include tools for background replacement or scene editing.
More questions about Kling AI? The Connector page answers them. See everything the Kling AI Connector can do
Connect Kling AI to Claude, Cursor, ChatGPT & more
