Topaz Labs/topazlabs-wonder-4-image/Released Nov 2026

Topaz Wonder 4

One-pass restoration, denoising, and upscaling

Commercial use

Image

About Topaz Wonder 4

Topaz Wonder 4 is a one-step image enhancement and upscaling model that repairs noise, blur, compression artifacts, image damage, and facial detail in a single pass while preserving natural textures.

A mixture-of-experts architecture identifies the regions of an image that need work and applies the right correction to each, so faces, fur, foliage, and fine textures are restored without a separate tuning pass. Compared with Wonder 3.5 it produces fewer artifacts and failures. Set the scale factor to upscale by up to 6x (the height scales to match); inputs and outputs can be up to 256 megapixels.

See what this model can do

A handful of real prompts and their outputs.

Side-by-side crop of a motorcycle V-twin engine, a smeared 280-pixel thumbnail on the left and crisp after 6x upscaling on the right.

Extreme Upscale

Engine from a Thumbnail

"Rebuilds pushrods, cooling fins, brass fittings and braided lines from a 280-pixel source."

Side-by-side crop of a jazz trumpeter mid-solo, grainy high-ISO noise on the left and clean detailed skin and brass on the right.

Low-Light Denoise

Noisy Club Portrait

"Strips heavy low-light grain while keeping skin texture, sweat and the shine on the brass."

Photo Restoration

1970s Family Picnic

"Brings back faces, plaid shirts and lakeside detail from a faded, soft family print."

Side-by-side crop of a corgi running on a beach, blocky with JPEG artifacts on the left and clean with fine fur on the right.

Compression Repair

Over-Shared Corgi

"Removes blocky JPEG artifacts from a twice-compressed photo and restores the fur."

Side-by-side crop of a blue dragonfly on a reed, out of focus on the left and sharp with veined wings on the right.

Deblur

Missed-Focus Dragonfly

"Recovers wing veins, legs and the reed from a shot where the focus missed."

Print Restoration

Seaside Postcard

"Sharpens the promenade crowd and the printed lettering of a faded postcard."

Ready to build with Topaz Wonder 4?

Try Topaz Wonder 4 in the Workbench to prompt it, compare outputs, and iterate on prompts without writing any code. When you're ready to ship, call the same model from our API and build your own apps on top of it.

curl -sSf -X POST https://hub.oxen.ai/api/ai/images/edit \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $OXEN_API_KEY" \
    -d '{
  "model": "topazlabs-wonder-4-image",
  "input_image": "https://hub.oxen.ai/api/repos/ox/Oxen-AI-Assets/file/main/images/ox_zoom_out_1926_1076.png",
  "scale": "1x"
}'

Request parameters

Every field Topaz Wonder 4 accepts in the request body. See the API reference for response formats.

POSThttps://hub.oxen.ai/api/ai/images/edit

Field	Type	Required	Default	Description
`input_image`	`string` uri	Yes	—	Image to enhance and upscale (URL or file path). jpeg, png, gif, or webp, up to 256 megapixels.
`scale`	`string`	No	`"1x"`	Upscale factor. 1x enhances at the original size; 2x, 4x, or 6x enlarge the image by that multiple (Topaz supports up to 6x and 256 megapixels). `1x2x4x6x`

Model JSON Image editing reference Async queue Models endpoint