Creative Image Models
Discover and call the current GPT Image, Seedream, Gemini, Grok, Kling, Krea, and Qwen image models
FocalAPI provides unified authentication for multiple creative image models while preserving each model's native parameter semantics. Use GET /v1/models to list models visible to the current key, then GET /v1/models/{model} for the model's live endpoint and parameter contract.
Availability, parameters, and pricing can vary by user group. Live model details and the console model catalog are authoritative. Do not use retired IDs or aliases that are absent from the list.
Unified image API
The following current models use POST /v1/images/generations. For editing, use reference-image fields or POST /v1/images/edits as shown by the selected model's details.
| Family | Current model IDs |
|---|---|
| GPT Image | gpt-image-2 |
| Seedream | seedream-4-0-250828, seedream-4-5-251128, dola-seedream-5-0-pro-260628, seedream-5-0-260128 |
| Grok Imagine | grok-imagine-image-quality, grok-imagine-image, grok-imagine-image-2.0 |
| Kling Image | kling-v3 |
| Krea 2 | krea-2-medium, krea-2-medium-turbo, krea-2-large |
| Qwen Image | qwen-image-3.0, qwen-image-3.0-pro |
curl https://api.focalapi.com/v1/images/generations \
-H "Authorization: Bearer $FOCALAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "seedream-5-0-260128",
"prompt": "A lighthouse by the sea in morning fog, cinematic lighting",
"size": "2k",
"watermark": false
}'Model-specific parameters
Do not apply one fixed parameter set to every model:
| Family | Important differences |
|---|---|
| GPT Image 2 | size (default 1024x1024, edges 1024–3840 in multiples of 16), quality (low/medium/high; omitted bills as low), background (auto/opaque), n 1–8, up to 16 reference images; mask requires exactly one reference image |
| Seedream | 4.0 defaults to 2048x2048 (tiers 1K/2K/4K, up to 10 reference images); 4.5 offers 2K/4K; 5.0 Pro generates a single image per request (tiers 1K/1.5K/2K); 5.0 (lite) accepts up to 14 reference images (tiers 2K/3K/4K); all versions add watermark and prompt optimization |
| Grok Imagine 2.0 | aspect_ratio, resolution, quality, and seed; up to 10 outputs |
| Kling Image 3.0 | aspect_ratio and negative_prompt; up to 9 outputs; a reference image cannot be combined with a negative prompt |
| Krea 2 | aspect_ratio, fixed 1K resolution, creativity, style references, and a moodboard; one output per request |
| Qwen Image 3.0 | size, up to three reference images, negative_prompt, prompt_extend, and watermark; up to six outputs |
Read exact constraints before constructing an advanced request:
curl https://api.focalapi.com/v1/models/kling-v3 \
-H "Authorization: Bearer $FOCALAPI_KEY"Native Gemini API
Gemini image models use the native generateContent endpoint:
| Current model ID | Product |
|---|---|
gemini-2.5-flash-image | At most 1 reference image, supplied as inlineData (fileData is not supported); fixed 1K output |
gemini-3-pro-image | Up to 14 reference images; imageSize 1K/2K/4K |
gemini-3.1-flash-image | Up to 14 reference images; imageSize 1K/2K/4K; 15 aspectRatio values (including 1:4, 4:1, 1:8, 8:1); supports thinkingLevel, temperature, and topP |
gemini-3.1-flash-lite-image | Up to 14 reference images; fixed 1K output; the same 15 aspectRatio values; supports thinkingLevel, temperature, and topP |
curl 'https://api.focalapi.com/v1beta/models/gemini-3.1-flash-image:generateContent' \
-H "Authorization: Bearer $FOCALAPI_KEY" \
-H "Content-Type: application/json" \
-d '{
"contents": [{"parts": [{"text": "A clean product illustration of a blue ceramic mug"}]}],
"generationConfig": {
"responseModalities": ["IMAGE", "TEXT"],
"responseFormat": {
"image": {"aspectRatio": "1:1", "imageSize": "1K"}
}
}
}'Images are returned as inlineData in response candidates. Whether aspectRatio, imageSize, and sampling fields are accepted still depends on the selected Gemini model's live details.
Billing
Pricing depends on the model and parameters such as output count, resolution, quality, and reference-image count. Check the console model catalog for current prices and discounts.