Overview
VS AI Render sends a VStitcher render to an AI image generation provider (Google AI Studio, OpenAI, or Stability AI — bring your own key) to produce a hyper-realistic product photo. Each provider receives the rendered garment image, your prompt and selected options, and your API key; a one-time consent dialog discloses this before the first Generate.
Provider | Model | Destination |
Google AI Studio | Gemini (image generation) |
|
OpenAI | GPT Image (images/edits) |
|
Stability AI | Stable Diffusion (structure control) |
|
Requirements
VStitcher 2025.3.0 or newer (BwApi 3.x), macOS or Windows
An API key for one of: Google AI Studio (Gemini image generation), OpenAI (GPT Image — requires Organization Verification on your OpenAI account before image generation works), or Stability AI (Stable Diffusion)
How to Generate a Render
Step 1: Configure a Provider
Open Plugins → Open VS AI Render → Settings, pick a provider, and paste that provider's API key. Only one key is stored at a time, tied to the selected provider — switching providers requires pasting that provider's key again.
Step 2: Set Generation Options
Option | Description |
Category | Male / Female / Kids |
Garment Type | Top / Bottom / Dress / Full Outfit |
Photoshoot Style | Studio, Streetwear, E-Commerce, Editorial |
Image Ratio | 2:3 (default), 1:1, 3:4, 4:3, 16:9, 9:16 |
Resolution | Provider-specific output resolution — see Resolution below. Not shown for Stability AI. |
Include Avatar | Include/exclude the avatar in the main render |
Additional Prompt | Optional free-text creative direction |
Step 3: Generate and Save
Click Generate Look, then Save Image (or Save All for multi-angle Ghost Product mode, written straight to ~/Desktop/Plugin Output/VSAIRender with no dialog). File format is PNG for OpenAI/Stability; Google AI Studio may return PNG, JPEG, or WebP depending on the model's response, and the saved file's extension matches whichever it was.
What Happens During Generation?
AI Vision Auto-Classify
Unless turned off, a separate preview render (always with avatar visible, for accurate body-shape detection) is sent to the provider first to auto-detect Category and Garment Type. The main render still respects the Include Avatar toggle either way.
Image Ratio
The selected ratio controls both the VStitcher render resolution and is included in the generation prompt as a composition instruction:
Ratio | Pixels |
2:3 Portrait (default) | 720 × 1080 |
1:1 Square | 1080 × 1080 |
3:4 Portrait | 810 × 1080 |
4:3 Landscape | 1080 × 810 |
16:9 Widescreen | 1920 × 1080 |
9:16 Tall | 1080 × 1920 |
Image Ratio only sizes the 3D render captured from VStitcher (the source photo sent to the AI) — it is not passed to any provider as an output resolution/size parameter. That's what Resolution (below) controls instead.
Resolution
Controls the AI provider's own output resolution, independent of Image Ratio:
Provider | Options |
Google AI Studio | 512px, 1K (default), 2K, 4K |
OpenAI | Auto (default), 1024×1024, 1024×1536, 1536×1024 |
Stability AI | Not shown — this provider's endpoint has no resolution setting; output always matches the input render's dimensions. |
Troubleshooting
Errors show as a categorized modal (Authentication / Connection / Rate Limit / Provider / Render) with a hint and a Copy Details button.
Issue | Fix |
Authentication Problem | Check the API key in Settings matches the selected provider |
"No open garment found" | Open a garment in VStitcher before generating |
Rate Limit Reached | Plugin retries automatically; check your quota with the provider if it persists |
AI Provider Error | The provider returned an unexpected response — try again, or switch providers in Settings |
No image returned (only text) | The model declined to generate an image for that prompt; falls back to the original VStitcher render |