ChatGPT-style virtual try-on API for your own store or app

Short answer: OpenAI added a “try on” button for clothing and accessories in ChatGPT (Retail Dive, 9 Oct 2026): a shopper uploads a photo and sees an image of themselves wearing the product. That button lives inside ChatGPT. To give your own shoppers or users the same thing, send a person photo and a garment image to TryOn-API.com — one OpenAI-compatible endpoint in front of dedicated try-on models (Kling Kolors) and generative image models (Google Gemini, OpenAI gpt-image-1, FAL FLUX Kontext).

How it works

1. Your app collects a person photo and the product (garment) image.
2. Your backend calls POST /api/v1/chat/completions with both images, tagged with tryon_role: "person" and tryon_role: "garment", or uploads them to the native multipart POST /api/v1/tryon endpoint.
3. You get back a generated image of the person wearing the garment and show it next to the product.

Code: the same OpenAI SDK you already use

Point any OpenAI-compatible SDK at TryOn-API.com — change the base URL, the key and the model slug:

// Any OpenAI-compatible SDK — change three things:
const client = new OpenAI({
  baseURL: "https://tryon-api.com/api/v1",   // 1. base URL
  apiKey:  process.env.TRYON_API_KEY,          // 2. your TryOn-API key
});
const res = await client.chat.completions.create({
  model: "kling/kolors-v1-5",                 // 3. a try-on model
  modalities: ["image"],
  messages: [{ role: "user", content: [
    { type: "text", text: "put the garment on the person" },
    { type: "image_url", image_url: { url: personDataUrl }, tryon_role: "person" },
    { type: "image_url", image_url: { url: garmentDataUrl }, tryon_role: "garment" },
  ]}],
});

Pick a model, switch any time

Every model sits behind the same schema, so you can compare try-on quality by changing one model slug and add a fallback chain with models[]. See the Models page for status and credits per request, and the clothing try-on rankings. For the trade-offs versus OpenRouter or calling providers directly, see Compare.

Frequently asked questions

Can I add ChatGPT-style try-on to my own store?

Yes. The ChatGPT try-on button works inside ChatGPT; to show try-on on your own product pages or in your own app, call a try-on API such as TryOn-API.com from your backend and display the generated image.

Is TryOn-API.com compatible with the OpenAI SDK?

Yes. POST /api/v1/chat/completions follows the OpenAI chat completions schema, including modalities:["image"] and streaming via SSE. Existing OpenAI SDKs work by changing the base URL, the API key and the model slug.

Which models can I use?

Dedicated try-on models such as Kling Kolors Virtual Try-On v1.5 and generative image models such as Google Gemini, OpenAI gpt-image-1 and FAL FLUX.1 Kontext, plus beta try-on models. The live list, with status and credits per request, is at GET /api/v1/models and on the Models page.

How is it priced?

Every request is billed in credits from one balance that works across all providers. Per-model prices are listed on the Models page: https://tryon-api.com/models

---

See also the machine-readable llms.txt and full API reference.