Will Standard V2 make a badly blurred face sharp?

In our defocused portrait it enlarged the blur. Wonder 4 reconstructed clearer eyes and facial detail, but generated detail should always be checked.

Are 2K, 4K and 8K fixed edge lengths?

These are approximate pixel-area tiers: 4, 16.8 and 67 MP. Our 1100×900 source returned at 2264×1852 on the 2K tier.

What changes the charge?

The preset and resolution tier determine the per-image charge. Output format does not affect it; 8K is the most expensive tier for each preset.

Topaz Gigapixel API — Photo Upscaling & Transparent PNG

topaz/gigapixel/upscale

Run one Gigapixel preset on a photo or graphic and receive an enlarged result in the selected pixel-area tier; transparent artwork is returned as PNG.

From $0.12/image

Run model View API

Input

Image upscaling

Input image1/1

Drop files or media links, paste, or browse · 1 file · Max 10 MB · JPG · PNG · WEBP

Style presetaffects price

Resolutionaffects price

Output format

JPEGPNG

Safety checkerOff

SpicyAPI does not run a safety checker on this model. Your prompt and media go to the model as written, and the model itself decides what comes back.

Checking your account before enabling submission and pricing.

Output

Ready for a new run

Example result

Edit

Split into layers

Animate

{
  "model": "topaz/gigapixel/upscale",
  "input": {
    "preset": "wonder_4",
    "resolution": "2k",
    "output_format": "jpeg",
    "image_url": "https://cdn.spicyapi.ai/models/examples/inputs/2c03c16cce72e642.jpg"
  }
}

Introduction

Gigapixel enlargement

Gigapixel offers faithful enlargement and generative reconstruction through a single preset choice. The default is standard_v2.

Request settings

Send image_url, choose preset and select resolution from 2k, 4k or 8k (default 4k). JPEG is the default output.

Delivery constraints

For transparent artwork use preset: transparent together with output_format: png. Review generated facial details before using the enlarged image.

Specifications for Topaz Gigapixel · Topaz Gigapixel
Field	Value
Input	Image
Output	Image
Source formats	JPEG, PNG, WebP
Default preset	`standard_v2`
Default tier	4K
Result formats	JPEG or PNG; transparent always PNG
Charging basis	Per image, by preset and output tier

Key features & highlights

More pixels alone do not fix missed focus

From left: source, Standard V2, Wonder 4. The precision result stays soft around the eyes; Wonder 4 rebuilds a more defined face while keeping its overall likeness in this test.

Check the person as well as the sharpness

Redefine made this man look younger and changed his features. Recover 3 and Standard MAX also rebuilt facial detail; Detail left this severely defocused face softer. These are observations from one image, not promises for every photo.

Keep the space around the rose transparent

Our cutout yellow-orange rose PNG grew from 400×301 to 2360×1776 with its alpha channel intact. Its pink petal tips carry water droplets, with a short green stem and leaves also visible. The dark comparison backdrop is for display only; the example file retains transparency.

Request parameters

Request parameters for Topaz Gigapixel · Topaz Gigapixel
Parameter	Required	Description
`image_url`	Yes	The single JPEG, PNG or WebP source that Gigapixel will enlarge. at least 1 characters
`preset`	No	Choose among twelve models. The five precision presets preserve content; generative presets reconstruct details. `standard_v2` · `high_fidelity_v3` · `low_resolution_v2` · `cgi` · `text_refine` · `wonder_4` · `wonder_3_5` · `recover_3` · `standard_max` · `redefine` · `detail` · `transparent` · Default: `standard_v2`
`resolution`	No	`2k`, `4k` (default) or `8k` selects approximately 4, 16.8 or 67 MP while retaining proportions. The tier also affects billing. `2k` · `4k` · `8k` · Default: `4k`
`output_format`	No	Gigapixel normally returns `jpeg`; `png` is also available at the same price. The `transparent` preset accepts only `png`. `jpeg` · `png` · Default: `jpeg`

Pricing

Gigapixel is billed per completed image according to preset and resolution. Changing output_format does not change the charge.

Pricing
Resolution	Style preset: cgi	Style preset: detail	Style preset: high_fidelity_v3	Style preset: low_resolution_v2	Style preset: standard_v2	Style preset: text_refine	Style preset: transparent	Style preset: wonder_3_5	Style preset: wonder_4	Style preset: recover_3	Style preset: redefine	Style preset: standard_max
`2k`	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.24 / image	$0.24 / image	$0.24 / image
`4k`	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.12 / image	$0.36 / image	$0.36 / image	$0.60 / image	$0.60 / image	$0.60 / image
`8k`	$0.36 / image	$0.36 / image	$0.36 / image	$0.36 / image	$0.36 / image	$0.36 / image	$0.36 / image	$1.08 / image	$1.08 / image	$2.04 / image	$2.04 / image	$2.04 / image

Use cases

Prepare an AI image for printing

Pick a resolution tier for a larger print, then inspect faces and fine patterns before sending the file to print.

Enlarge a logo or a cutout

Try text_refine for graphics and transparent for a PNG whose background must remain clear.

Prompt guide & pro tips

Supply the source

Set image_url to your JPEG, PNG or WebP photo or artwork.
Decide how much to rebuild

Start with standard_v2 for faithful enlargement or select wonder_4 when lost detail needs reconstruction.
Choose the delivery size

Set resolution and output_format; for transparent, explicitly request png.

Gigapixel enlargement

Run one Gigapixel preset on a photo or graphic and receive an enlarged result in the selected pixel-area tier; transparent artwork is returned as PNG.

Request settings

{
  "image_url": "https://example.com/blurred-portrait.jpg",
  "preset": "standard_v2",
  "resolution": "2k",
  "output_format": "jpeg"
}

Preset guide

`preset`	Best starting point
`standard_v2`	General photos; default precision option
`high_fidelity_v3`	Already sharp, high-resolution material
`low_resolution_v2`	Tiny or heavily compressed images
`cgi`	Illustrations and rendered artwork
`text_refine`	Text, logos and graphics
`wonder_4`	Latest all-purpose generative enhancement
`wonder_3_5`	Generative reconstruction with Wonder 3.5
`recover_3`	Severely degraded photographs
`standard_max`	Generative detail reconstruction
`redefine`	Adds new detail and may change likeness
`detail`	More realistic skin in AI-generated faces
`transparent`	Alpha-preserving PNG enlargement

Observed results

From left: source, Standard V2, Wonder 4. The precision result stays soft around the eyes; Wonder 4 rebuilds a more defined face while keeping its overall likeness in this test.

Delivery constraints

A request handles one image with one preset. Generative presets can alter faces and small details, so they are unsuitable for identity documents or evidentiary restoration. An 8K result can be a large download. Choosing JPEG with transparent is rejected.

API quick start

# Generate the idempotency key once; reuse it when retrying after a lost response.
export SPICY_IDEMPOTENCY_KEY="${SPICY_IDEMPOTENCY_KEY:-$(uuidgen)}"
curl -X POST https://api.spicyapi.ai/api/v1/jobs/createTask \
  -H "Authorization: Bearer $SPICY_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $SPICY_IDEMPOTENCY_KEY" \
  -d '{
  "model": "topaz/gigapixel/upscale",
  "input": {
    "image_url": "https://example.com/reference.png",
    "output_format": "jpeg",
    "preset": "standard_v2",
    "resolution": "4k"
  }
}'

# Copy data.taskId from the response into TASK_ID, then poll the status.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.spicyapi.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
  -H "Authorization: Bearer $SPICY_API_KEY"

See the API reference for authentication, the complete schema, polling, and error handling.

Related models and next tasks

Missing the model you need?

TOPAZ GIGAPIXEL API

Make a small image usable at print size, with a choice over how much gets rebuilt.

Key numbers

Preset choices: 12; Five precision models, six generative models and one transparent-image preset.
Output tiers: 2K–8K; Approximately 4, 16.8 or 67 megapixels; the original aspect ratio stays.
Alpha preserved: PNG; The transparent preset requires PNG output, even when JPEG is your usual format.

Before you spend

Upscaling

What it does

A blurred portrait

More pixels alone do not fix missed focus

From left: source, Standard V2, Wonder 4. The precision result stays soft around the eyes; Wonder 4 rebuilds a more defined face while keeping its overall likeness in this test.

Creative reconstruction

Check the person as well as the sharpness

Cutout rose

Keep the space around the rose transparent

Where it fits

Prepare an AI image for printing
Pick a resolution tier for a larger print, then inspect faces and fine patterns before sending the file to print.
Enlarge a logo or a cutout
Try `text_refine` for graphics and `transparent` for a PNG whose background must remain clear.

Made with Topaz Gigapixel

Model reference

Topaz Gigapixel API reference

Topaz Gigapixel API specifications

Source formats: JPEG, PNG, WebP
Default preset: `standard_v2`
Default tier: 4K
Result formats: JPEG or PNG; transparent always PNG
Charging basis: Per image, by preset and output tier

How to call the Topaz Gigapixel API

Supply the source
Set `image_url` to your JPEG, PNG or WebP photo or artwork.
Decide how much to rebuild
Start with `standard_v2` for faithful enlargement or select `wonder_4` when lost detail needs reconstruction.
Choose the delivery size
Set `resolution` and `output_format`; for `transparent`, explicitly request `png`.

The short version

For Gigapixel, choose the reconstruction approach before choosing the output tier: a faithful enlargement and a newly detailed face solve different problems.

Workbench

Input

Output

Introduction

Gigapixel enlargement

Request settings

Delivery constraints

Key features & highlights

More pixels alone do not fix missed focus

Check the person as well as the sharpness

Keep the space around the rose transparent

Request parameters

Pricing

Use cases

Prepare an AI image for printing

Enlarge a logo or a cutout

Prompt guide & pro tips

Gigapixel enlargement

Request settings

Preset guide

Observed results

Delivery constraints

API quick start

Related models and next tasks

Make a small image usable at print size, with a choice over how much gets rebuilt.

Key numbers

Before you spend

What it does

More pixels alone do not fix missed focus

Check the person as well as the sharpness

Keep the space around the rose transparent

Where it fits

Prepare an AI image for printing

Enlarge a logo or a cutout

Made with Topaz Gigapixel

Topaz Gigapixel API reference

Topaz Gigapixel API specifications

How to call the Topaz Gigapixel API

Supply the source

Decide how much to rebuild

Choose the delivery size

The short version

Topaz Gigapixel API questions

Will Standard V2 make a badly blurred face sharp?+

Are 2K, 4K and 8K fixed edge lengths?+

What changes the charge?+

All Topaz models6

Image models3

Video models3

Related comparisons