Skip to content

Gigapixel High Fidelity 3

Image-to-Image • topaz

View as MarkdownAgent setup

Model ID

Use with the Workers AI binding, the AI Gateway REST API, or the Unified API.

Calling the provider directly? Use the provider's model name instead, including with provider-native endpoints or the provider's own SDK. For example, Anthropic's name for anthropic/claude-opus-4.8 is claude-opus-4-8.

  • Third-party

Topaz Labs Gigapixel High Fidelity 3 is a precision image upscaler tuned for high-resolution sources and professional photography, preserving fine texture and detail with minimal alteration. It adds recovery-strength and opacity controls over High Fidelity 2.

Model Info
More informationlink ↗
Pricing
  • Per provider unit$0.12

Usage

const response = await env.AI.run(
  'topaz/gigapixel-high-fidelity-3',
  {
    image:
      'https://images-assets.nasa.gov/image/GSFC_20171208_Archive_e000842/GSFC_20171208_Archive_e000842~medium.jpg',
    output_format: 'jpeg',
    output_width: 2560,
  },
)
console.log(response)
curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run \
  --header "Authorization: Bearer $CLOUDFLARE_API_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "topaz/gigapixel-high-fidelity-3",
  "input": {
    "image": "https://images-assets.nasa.gov/image/GSFC_20171208_Archive_e000842/GSFC_20171208_Archive_e000842~medium.jpg",
    "output_format": "jpeg",
    "output_width": 2560
  }
}'
Star Field Upscale
{
  "state": "Completed",
  "result": {
    "image": "https://examples.aig.cloudflare.com/topaz/gigapixel-high-fidelity-3/star-field-upscale.jpeg"
  }
}

Parameters

image
stringrequiredSource image as a public HTTPS URL or a base64 data URI (JPEG, PNG, or TIFF)
output_width
integerminimum: 1maximum: 32000Desired output width in pixels. If only one dimension is set, the other scales proportionally. If neither is set, the source resolution is kept (enhancement without upscaling)
output_height
integerminimum: 1maximum: 32000Desired output height in pixels. If only one dimension is set, the other scales proportionally. If neither is set, the source resolution is kept (enhancement without upscaling)
output_format
stringrequireddefault: jpegenum: jpeg, png, tiffFormat of the upscaled image
crop_to_fill
booleanCrop to fill the requested dimensions instead of letterboxing when the aspect ratio differs
face_enhancement
booleanRun face recovery on detected faces
face_enhancement_strength
numberminimum: 0maximum: 1Strength of face recovery
face_enhancement_creativity
numberminimum: 0maximum: 1Face recovery style, from realistic (0) to creative (1)
subject_detection
stringenum: foreground, background, allWhere enhancements are applied
sharpen
numberminimum: 0maximum: 1Amount of additional sharpening
denoise
numberminimum: 0maximum: 1Amount of noise reduction
fix_compression
numberminimum: 0maximum: 1Amount of compression-artifact reduction
recovery_strength
numberminimum: 0maximum: 1Strength of detail recovery (Topaz default 1.0)
opacity
numberminimum: 0maximum: 1Blend of the enhanced result over the source (Topaz default 1.0)
image
stringformat: uriTime-limited presigned URL to the upscaled image

API Schemas (Raw)

Input
Output

Was this helpful?