GET

/v1/schemas/fal/image/fal-ai/image2pixel

kind inputactivity imagehash 8b56ad5d…extracted Jul 6, 2026version currentsource upstream spec ↗
fieldrecorded source
schemaSource linked above.
Source trace
570244531544bbe3689370d79c8ab7cb54c3cdc1e86f676aa2da65a4f2003191
{
  "fetchedAt": 1783330412,
  "extractorVersion": "1",
  "verifiedAt": null
}

Request body

17 properties · 1 required · 0 $defs
propertytypedescriptionconstraints
image_url*stringThe image URL to process into improved pixel art
dominant_color_thresholdnumberDominant color threshold (0.0-1.0).0 ≤ n ≤ 1 · default: 0.05
trim_bordersbooleanTrim borders of the image.default: false
detect_methodenumScale detection method to use.auto | runs | edge · default: "auto"
downscale_methodenumDownscaling method to produce the pixel-art output.dominant | median | mode | mean | content-adaptive · default: "dominant"
sync_modebooleanIf `True`, the media will be returned as a data URI and the output data won't be available in the request history.default: false
fixed_palettestring[]Optional fixed color palette as hex strings (e.g., ['#000000', '#ffffff']).nullable
cleanup_morphbooleanApply morphological operations to remove noise.default: false
alpha_thresholdintegerAlpha binarization threshold (0-255).0 ≤ n ≤ 255 · default: 128
scaleintegerForce a specific pixel scale. If None, auto-detect.1 ≤ n ≤ 64 · nullable
auto_color_detectbooleanEnable automatic detection of optimal number of colors.default: false
snap_gridbooleanAlign output to the pixel grid.default: true
cleanup_jaggybooleanRemove isolated diagonal pixels (jaggy edge cleanup).default: false
max_colorsintegerMaximum number of colors in the output palette. Set None to disable limit.1 ≤ n ≤ 256 · nullable
background_toleranceintegerBackground tolerance (0-255).0 ≤ n ≤ 255 · default: 0
background_modeenumControls where to flood-fill from when removing the background.edges | corners | midpoints · default: "corners"
transparent_backgroundbooleanRemove background of the image. This will check for contiguous color regions from the edges after correction and make them transparent.default: false

Validate a payload

before you spend tokens
shell
$ curl -X POST https://modelschemas.com/v1/validate \
    -d '{"provider":"fal","endpointId":"fal-ai/image2pixel","payload":{…}}'
→ {"valid": false, "errors": [{"path": "#", "keyword": "required", …}]}