Provider Topaz

Topaz

Mode: image · Models: 9

Vendor: Topaz Labs · Official API docs: developer.topazlabs.com

Topaz Labs provides AI image enhancement and upscaling. The flagship topaz-upscale-image does a 2× HD upscale with face enhancement preset; the remaining variants are enhancement models tuned for different source types — low-res photos, CGI/3D renders, high-fidelity preservation, text/graphics, and generative recovery. All variants run through the topaz/upscale/image workflow and take an image as input.

Models

id	Name	Input type
topaz-upscale-image	Topaz HD Upscale	i2i
topaz-enhance-standard-v2	Topaz Standard	i2i
topaz-enhance-low-res-v2	Topaz Low Res	i2i
topaz-enhance-cgi	Topaz CGI	i2i
topaz-enhance-high-fidelity-v2	Topaz Hi-Fi	i2i
topaz-enhance-text-refine	Topaz Text	i2i
topaz-enhance-redefine	Topaz Redefine	i2i
topaz-enhance-recovery	Topaz Recovery	i2i
topaz-enhance-recovery-v2	Topaz Recovery V2	i2i

CLI

# 2× HD upscale of a photo
gen-ai generate -m topaz-upscale-image -i ./portrait.jpg

# enhance a low-resolution source
gen-ai generate -m topaz-enhance-low-res-v2 -i ./old-scan.jpg

# refine text and graphics
gen-ai generate -m topaz-enhance-text-refine -i ./poster.png

MCP

{ "name": "picsart_generate",
  "arguments": {
    "model": "topaz-upscale-image",
    "imageUrls": ["https://example.com/portrait.jpg"]
  } }

The task-shaped picsart_enhance tool also routes to Topaz upscaling:

{ "name": "picsart_enhance",
  "arguments": {
    "imageUrls": ["https://example.com/portrait.jpg"]
  } }

Parameters — topaz-upscale-image

Param	CLI flag	Type	Values
imageUrls	-i	file	required · single image (max 1)

Source: gen-ai models info topaz-upscale-image --json. The upscale model exposes only the input image; Topaz's automatic settings handle the rest (the catalog presets upscale_factor: 2, model: Standard V2, face enhancement, and output_format: png). The enhance variants share this surface, differing only in the enhancement model they send. See the Topaz API reference for the underlying /enhance parameters (output dimensions, crop_to_fill, face_enhancement, subject_detection).

Pricing

gen-ai pricing topaz-upscale-image

Topaz bills per output megapixels (roughly 1 credit per 24 MP of output), so cost scales with the upscaled resolution rather than per-image flat. All variants share the topaz-upscale-image pricing key.