← All guides

Guide

Upscale a whole catalog of product photos at once

November 2026 · 4 minute read

Old product photos are often too small: 500 pixels wide from a supplier, a thumbnail from an old site, or a phone shot cropped down. Marketplaces and themes want bigger images, and zoom often only works above a minimum size (check your marketplace’s current image rules). Plain resizing makes small photos soft. An AI upscaler rebuilds sharp edges instead. This guide upscales a whole list of photos in one run and downloads the results into a folder.

This guide uses our AI Image Upscaler on the Apify Store. Every step works with a free Apify account.

What you’ll end up with

A folder of upscaled files, named in order (upscaled-0001.jpg, upscaled-0002.jpg …), and a table that maps each original link to its new file:

imageUrlstatusinputWidthoutputWidthoutputKey
https://…/mug-500.jpgupscaled5002000upscaled-0001.jpg
https://…/old-page.htmlfailed

Failed rows say why (for example, the link opens a web page, not an image) and cost nothing.

1. Collect direct image links

The Actor downloads each image from a link, so you need links that open the image file itself:

Links to a web page, a gallery or a Google Drive preview page won’t work: those open HTML, not the image.

2. Pick the settings

3. Run it

  1. Open the AI Image Upscaler on Apify (its Apify Store page) and click Try for free. A free Apify account comes with monthly credit.
  2. Paste your links into Image URLs, set the options above and click Start.
  3. Open the Results table. Each row links to its upscaled file.

Time and cost: a 500 × 500 photo takes about 5–20 seconds; a 1920 × 1080 one about 3 minutes at the default memory. You pay per image, by the size you send in, and failed images are free. The Actor’s Pricing tab has the price per size. Set a maximum cost in the run options if you want a hard cap.

4. Download everything into a folder

For a handful of images, click each outputUrl in the Results table. For hundreds, run this Python script once (install the client with pip install apify-client, and copy your API token from Apify Console → Settings → API & Integrations):

from pathlib import Path
from apify_client import ApifyClient

client = ApifyClient("<YOUR_API_TOKEN>")
run = client.actor("clearpull/image-upscaler").call(run_input={
    "imageUrls": [line.strip() for line in open("links.txt") if line.strip()],
    "scale": "2",
    "imageType": "photo",
    "outputFormat": "jpeg",
})

store = client.key_value_store(run["defaultKeyValueStoreId"])
out = Path("upscaled")
out.mkdir(exist_ok=True)
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    if row["status"] != "upscaled":
        print("skipped", row["imageUrl"], row.get("error"))
        continue
    record = store.get_record_as_bytes(row["outputKey"])
    (out / row["outputKey"]).write_bytes(record["value"])
    print(row["imageUrl"], "->", row["outputKey"], f'{row["outputWidth"]}x{row["outputHeight"]}')

links.txt holds one image link per line. The script prints which original became which file, so you can rename them or match them back to products.

Tips

The models are Real-ESRGAN’s realesr-general-x4v3 and realesr-animevideov3 (BSD 3-Clause), run on CPUs, so there’s no GPU to rent or set up.

Need data from another public source, or a different setup? Get in touch.