Images
POST /v1/image — generate images through the assigned route.
Request
curl https://ai.latentkit.com/v1/image \
-H "Authorization: Bearer $LATENTKIT_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"prompt": "A clean product icon on a white background",
"size": "1024x1024"
}'Fields
| Field | Description |
|---|---|
prompt | Image description (required) |
size | Output size when supported by the route model (e.g. 1024x1024) |
quality | Quality tier when supported (e.g. standard, hd) |
style | Style hint when supported (e.g. vivid, natural) |
n | Number of images when supported |
response_format | Output format hint when supported (e.g. url, b64_json) |
response_profile | fast, balanced, or deep when the route allows overrides |
The route must include a model with the image_generation capability (for example DALL·E or Grok Imagine image models). Unsupported option values are ignored or rejected by the upstream model.
SDK
const image = await client.image.generate({
prompt: 'A clean product icon on a white background',
size: '1024x1024',
});image = client.image.generate(
prompt="A clean product icon on a white background",
size="1024x1024",
)Responses follow the unified runtime shape for your route's provider adapter. Inspect the returned payload for URLs or base64 content fields.
Billing
Platform Access image generation consumes workspace credits. BYOK image generation is charged by your provider account. Check current model pricing in the console.
Free plan note
Plain image generation requests may count toward a workspace managed onboarding allowance on Free plans. After limits or without credits, add credits or use BYOK routes.
Related
- Vision — analyzing images instead of generating them
- Video
- Models & providers