Available models
Loading…
See text-to-image models in the Model Catalog and filter for Image. The table lists supported version identifiers; actual availability follows the live catalog for the current environment. SeeDream is billed per image.
Request parameters
Header parameters
AuthorizationstringRequiredAPI key in the format Bearer <key>.
Example:
Bearer sk-zerofa-xxxContent-TypestringRequiredExample:
application/jsonBody parametersapplication/json
modelstringRequiredA SeeDream model. See available models above.
Example:
your-seedream-model-idpromptstringRequiredA detailed image description. Chinese prompts are supported.
Example:
A shiba inu wearing a space helmet, flat illustration, studio lightnintegerOptionalNumber of images to generate, default 1.
Example:
1sizestringOptionalOutput resolution. Note the minimum pixel requirement in Volcano-specific behavior below.
Example:
2048x2048Volcano does not support response_format. SeeDream always returns image URLs.
Volcano-specific behavior
- Minimum output is approximately 1920×1920 pixels. Smaller sizes are scaled proportionally, so output may be larger and slower than requested. Use a valid size such as 2048x2048 for precise control.
- Returns URLs rather than base64. Links are temporary signed Volcano TOS URLs that usually expire after about 24 hours, so download them promptly.
- Two image-to-image paths are available: /v1/images/generations with an image URL is recommended, while OpenAI-compatible /v1/images/edits accepts uploaded files.
- Billed by actual image count rather than tokens. Prices vary by model and size.
Native image-to-image (recommended)
Add an image URL to /v1/images/generations for native image-to-image without multipart upload.
Additional body parameters for native Volcano i2i
imagestring | string[]OptionalReference images as a URL, base64 data URI, or array. SeeDream supports 1–14 images; including image enables image-to-image mode.
Example:
https://...pngsequential_image_generationstringOptionaldisabled produces one image by default; auto can generate a sequence of images.
Example:
disabledsequential_image_generation_optionsobjectOptionalSequence settings such as max_images. auto mode is billed by the actual number returned.
Example:
{"max_images":3}Native image-to-image · curl (image URL)
curl https://zerofa.ai/v1/images/generations \
-H "Authorization: Bearer sk-zerofa-xxx" \
-H "Content-Type: application/json" \
-d '{
"model": "your-seedream-model-id",
"prompt": "Replace the background with a starry night sky while preserving the subject",
"image": "https://your-cdn.com/cat.png",
"size": "2048x2048"
}'Image sequence · sequential=auto
{
"model": "your-seedream-model-id",
"prompt": "Create three variations with different poses from this image",
"image": ["https://your-cdn.com/cat.png"],
"sequential_image_generation": "auto",
"sequential_image_generation_options": { "max_images": 3 }
}The image parameter is specific to Volcano SeeDream. Use /v1/images/edits for image-to-image with other providers.
OpenAI-compatible image-to-image
POST
/v1/images/editsUpload an image and prompt with multipart/form-data. ZeroFA converts the file to base64 and calls SeeDream generations, fixed to one output. Use the native format above for URLs or image sequences.
Image edit · curl (file upload)
curl https://zerofa.ai/v1/images/edits \
-H "Authorization: Bearer sk-zerofa-xxx" \
-F "model=your-seedream-model-id" \
-F "prompt=Replace the background with a starry night sky" \
-F "size=2048x2048" \
-F "image=@source.png;type=image/png"Response
Each data item is one image with a temporary Volcano TOS URL. Download it for long-term storage.
Request and response bodies
Use the examples below to verify the request and response structures. To send a request, select Debug at the top of the page.
curl https://zerofa.ai/v1/images/generations \
-H "Authorization: Bearer sk-zerofa-xxx" \
-H "Content-Type: application/json" \
-d '{
"model": "your-seedream-model-id",
"prompt": "A shiba inu wearing a space helmet, flat illustration, studio light",
"n": 1,
"size": "2048x2048"
}'Response · 200
{
"created": 1780943800,
"data": [
{ "url": "https://upstream-cdn.example.com/generated/image.jpeg?..." }
]
}