Docs/SparkPix API

SparkPix API

SparkPix's ultra-fast image models with two separate endpoints. SparkPix Image generates text-to-image in under 1 second with LoRA custom style support. SparkPix Image Edit supports multi-image reference editing. Both are synchronous APIs returning results directly.

Authentication

Authorization: Bearer YOUR_API_KEY

Models & pricing

Model	Endpoint	Price	Notes
SparkPix Image	POST /api/v1/images/generations-sync	$0.0100/image	Text-to-image, sync, <1s, LoRA support
SparkPix Image Edit	POST /api/v1/images/edit	$0.0130/image	Multi-image editing, sync, <1s

Code examples

# SparkPix Image: Text-to-image (sync, sub-1-second)
curl -X POST https://apimodels.app/api/v1/images/generations-sync \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "A cute cat wearing a space helmet, digital art",
    "aspect_ratio": "16:9",
    "seed": 42
  }'

# SparkPix Image Edit: Multi-image editing (sync)
curl -X POST https://apimodels.app/api/v1/images/edit \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Make the background a sunset beach",
    "images": ["https://example.com/photo.jpg"],
    "aspect_ratio": "16:9"
  }'

SparkPix Image — Request parameters

POST /api/v1/images/generations-sync

Field	Required	Type	Description
prompt	Yes	string	Text description for image generation
aspect_ratio	No	string	"custom" (default, auto size), "1:1", "16:9", "9:16", "4:3", "3:4", "3:2", "2:3"
width	No	integer	Only when aspect_ratio="custom", 256-1440, must be multiple of 16
height	No	integer	Only when aspect_ratio="custom", 256-1440, must be multiple of 16
prompt_upsampling	No	boolean	Auto-enhance prompt with LLM (default: false)
seed	No	integer	Random seed for reproducibility
lora_weights	No	string	HuggingFace LoRA weights URL, format: huggingface.co/<owner>/<model>[/<file.safetensors>]
lora_scale	No	number	LoRA strength, -1 to 3, default 0.5
callBackUrl	No	string	Webhook URL called when task completes

SparkPix Image Edit — Request parameters

POST /api/v1/images/edit

Field	Required	Type	Description
prompt	Yes	string	Editing instructions. Refer to inputs as "image 1", "image 2", etc.
images	No	string[]	Reference image URL array, first image is the primary one
aspect_ratio	No	string	"match_input_image" (default, match first image), "1:1", "16:9", "9:16", "4:3", "3:4", "3:2", "2:3"
turbo	No	boolean	Turbo mode (default: true), recommended off for complex tasks
seed	No	integer	Random seed for reproducibility
callBackUrl	No	string	Webhook URL called when task completes

Notes

The two models use different API endpoints
Both are synchronous APIs returning image URLs directly, no polling needed
Response format: { code: 200, data: { url, model, credits, duration } }
SparkPix Image aspect_ratio defaults to "custom" (auto), SparkPix Image Edit defaults to "match_input_image"

Try it in the Playground

FAQ

How fast is SparkPix?

SparkPix Image generates text-to-image in under 1 second and both endpoints are synchronous — the image URL comes back in the same response, no polling.

Which endpoint do I use for editing vs generation?

Text-to-image: POST /api/v1/images/generations-sync. Editing: POST /api/v1/images/edit with images[] (first image is the primary). They are separate endpoints.