Upscaling Images

We’re going to walk through how to take a small, low-resolution image and upscale it to a higher resolution using the upscale endpoint. This is useful for sharpening generated thumbnails, restoring detail in old photos, or producing print-ready assets.

We’ll be using a 384x384 lighthouse photo as our input image:

Project Setup

# Create a project directory.
mkdir prodia-upscaling-images
cd prodia-upscaling-images

prodia-js

Install Node (if not already installed):

macOS

brew install node
# Close the current terminal and open a new one so that node is available.

Linux

apt install node
# Close the current terminal and open a new one so that node is available.

Windows

winget install -e --id OpenJS.NodeJS.LTS
# Close the current terminal and open a new one so that node is available.

Create project skeleton:

# Requires node --version >= 18
# Initialize the project with npm.
npm init -y

# Install the prodia-js library.
npm install prodia --save

requests

Install Python (if not already installed):

macOS

brew install python
# Close the current terminal and open a new one so that python is available.

Linux

apt install python3 python3-venv python-is-python3
# Close the current terminal and open a new one so that python is available.

Windows

winget install -e --id Python.Python.3.12
# Close the current terminal and open a new one so that python is available.

# Requires python --version >= 3.12
python -m venv venv
source venv/bin/activate
pip install requests

curl

Install curl (if not already installed):

macOS

brew install curl
# Close the current terminal and open a new one so that curl is available.

Linux

apt install curl
# Close the current terminal and open a new one so that curl is available.

Windows

# NOTE: Windows 10 and up have curl installed by default and this can be
# skipped.
winget install -e --id cURL.cURL
# Close the current terminal and open a new one so that curl is available.

# Export your token so it can be used by the main code.
export PRODIA_TOKEN=your-token-here

Your token is exported to an environment variable. If you close or switch your shell you’ll need to run export PRODIA_TOKEN=your-token-here again.

Create a main file for your project:

prodia-js

main.js

const { createProdia } = require("prodia/v2");

const prodia = createProdia({
    token: process.env.PRODIA_TOKEN // get it from environment
});

requests

Create the following main.py

main.py

from requests.adapters import HTTPAdapter, Retry
import os
import requests
import sys


prodia_token = os.getenv('PRODIA_TOKEN')
prodia_url = 'https://inference.prodia.com/v2/job'

session = requests.Session()
retries = Retry(allowed_methods=None, status_forcelist=Retry.RETRY_AFTER_STATUS_CODES)
session.mount('http://', HTTPAdapter(max_retries=retries))
session.mount('https://', HTTPAdapter(max_retries=retries))
session.headers.update({'Authorization': f"Bearer {prodia_token}"})

curl

main.sh

set -euo pipefail

You’re now ready to make some API calls!

Upscale 2x

The upscale endpoint accepts a single image and returns a larger version of it. The upscale config field controls the factor — 2, 4, or 8.

prodia-js

main.js

const { createProdia } = require("prodia/v2");
const fs = require("node:fs/promises");

const prodia = createProdia({
  token: process.env.PRODIA_TOKEN,
});

(async () => {
  // get input image
  const inputBuffer = await (await fetch("https://docs.prodia.com/upscale-input.jpg")).arrayBuffer();

  const job = await prodia.job({
    type: "inference.upscale.v1",
    config: {
      upscale: 2,
    },
  }, {
    inputs: [ inputBuffer ]
  });

  const image = await job.arrayBuffer();
  await fs.writeFile("upscaled.jpg", new Uint8Array(image));
})();

node main.js

requests

main.py

from requests.adapters import HTTPAdapter, Retry
from io import BytesIO
import json
import os
import requests
import sys


prodia_token = os.getenv('PRODIA_TOKEN')
prodia_url = 'https://inference.prodia.com/v2/job'

session = requests.Session()
retries = Retry(allowed_methods=None, status_forcelist=Retry.RETRY_AFTER_STATUS_CODES)
session.mount('http://', HTTPAdapter(max_retries=retries))
session.mount('https://', HTTPAdapter(max_retries=retries))
session.headers.update({'Authorization': f"Bearer {prodia_token}"})

try:
    with open('upscale-input.jpg', 'rb') as f:
        input_image = f.read()
except FileNotFoundError:
    res = requests.get('https://docs.prodia.com/upscale-input.jpg')
    input_image = res.content
    with open('upscale-input.jpg', 'wb') as f:
        f.write(res.content)

headers = {
    'Accept': 'image/jpeg',
}

job = {
    'type': 'inference.upscale.v1',
    'config': {
        'upscale': 2,
    },
}

files = [
    ('job', ('job.json', BytesIO(json.dumps(job).encode('utf-8')), 'application/json')),
    ('input', ('upscale-input.jpg', input_image, 'image/jpeg')),
]

res = session.post(prodia_url, headers=headers, files=files)
print(f"Request ID: {res.headers['x-request-id']}")
print(f"Status: {res.status_code}")

if res.status_code != 200:
    print(res.text)
    sys.exit(1)

with open('upscaled.jpg', 'wb') as f:
    f.write(res.content)

python main.py

curl

main.sh

set -euo pipefail

cat <<EOF > job.json
{
  "type": "inference.upscale.v1",
  "config": {
    "upscale": 2
  }
}
EOF

if [[ ! -f upscale-input.jpg ]]; then
  curl -Lo upscale-input.jpg 'https://docs.prodia.com/upscale-input.jpg'
fi

curl -sSf --retry 3 \
  -H "Authorization: Bearer $PRODIA_TOKEN" \
  -H 'Accept: image/jpeg' \
  -F [email protected] \
  -F [email protected] \
  --output upscaled.jpg \
  https://inference.prodia.com/v2/job

bash main.sh

macOS

open upscaled.jpg

Linux

xdg-open upscaled.jpg

Windows

start upscaled.jpg

The 384x384 input is now a sharp 768x768 image. Edges around the lighthouse and rocks have crisper detail and the JPEG compression artifacts in the sky are gone:

Upscale 4x

For larger outputs, set upscale to 4 or 8. The endpoint returns the same dimensions multiplied by the chosen factor — for our 384x384 input, upscale: 4 produces a 1536x1536 image.

prodia-js

main.js

const { createProdia } = require("prodia/v2");
const fs = require("node:fs/promises");

const prodia = createProdia({
  token: process.env.PRODIA_TOKEN,
});

(async () => {
  // get input image
  const inputBuffer = await (await fetch("https://docs.prodia.com/upscale-input.jpg")).arrayBuffer();

  const job = await prodia.job({
    type: "inference.upscale.v1",
    config: {
      upscale: 4,
    },
  }, {
    inputs: [ inputBuffer ]
  });

  const image = await job.arrayBuffer();
  await fs.writeFile("upscaled-4x.jpg", new Uint8Array(image));
})();

node main.js

requests

main.py

from requests.adapters import HTTPAdapter, Retry
from io import BytesIO
import json
import os
import requests
import sys


prodia_token = os.getenv('PRODIA_TOKEN')
prodia_url = 'https://inference.prodia.com/v2/job'

session = requests.Session()
retries = Retry(allowed_methods=None, status_forcelist=Retry.RETRY_AFTER_STATUS_CODES)
session.mount('http://', HTTPAdapter(max_retries=retries))
session.mount('https://', HTTPAdapter(max_retries=retries))
session.headers.update({'Authorization': f"Bearer {prodia_token}"})

try:
    with open('upscale-input.jpg', 'rb') as f:
        input_image = f.read()
except FileNotFoundError:
    res = requests.get('https://docs.prodia.com/upscale-input.jpg')
    input_image = res.content
    with open('upscale-input.jpg', 'wb') as f:
        f.write(res.content)

headers = {
    'Accept': 'image/jpeg',
}

job = {
    'type': 'inference.upscale.v1',
    'config': {
        'upscale': 4,
    },
}

files = [
    ('job', ('job.json', BytesIO(json.dumps(job).encode('utf-8')), 'application/json')),
    ('input', ('upscale-input.jpg', input_image, 'image/jpeg')),
]

res = session.post(prodia_url, headers=headers, files=files)
print(f"Request ID: {res.headers['x-request-id']}")
print(f"Status: {res.status_code}")

if res.status_code != 200:
    print(res.text)
    sys.exit(1)

with open('upscaled-4x.jpg', 'wb') as f:
    f.write(res.content)

python main.py

curl

main.sh

set -euo pipefail

cat <<EOF > job.json
{
  "type": "inference.upscale.v1",
  "config": {
    "upscale": 4
  }
}
EOF

if [[ ! -f upscale-input.jpg ]]; then
  curl -Lo upscale-input.jpg 'https://docs.prodia.com/upscale-input.jpg'
fi

curl -sSf --retry 3 \
  -H "Authorization: Bearer $PRODIA_TOKEN" \
  -H 'Accept: image/jpeg' \
  -F [email protected] \
  -F [email protected] \
  --output upscaled-4x.jpg \
  https://inference.prodia.com/v2/job

bash main.sh

macOS

open upscaled-4x.jpg

Linux

xdg-open upscaled-4x.jpg

Windows

start upscaled-4x.jpg

At 4x, the rocks gain individual texture and the clouds resolve into distinct shapes:

Parameters

Parameter	Type	Values	Default	Description
upscale	number	2, 4, 8	2	How many times larger the output should be along each dimension.

Picking a factor

Goal	Factor
Sharpen a generated image without changing dimensions much	2
Produce a print-ready or hero-banner asset from a thumbnail	4
Maximum resolution for archival or large-format display	8

Outputs scale quadratically — an 8x upscale of a 1024x1024 input produces an 8192x8192 image (around 67 megapixels), so prefer the smallest factor that meets your needs.

See also

Restoring Faces If your input contains faces, run face restoration first or use the combined facerestore.upscale endpoint.

Generating Images Generate a base image at lower resolution, then upscale to save time and cost.