Website Images¶

Download images from a web page and convert them to a chosen file format.

Overview¶

Website Images scans a website for common image references, follows same-site page links, downloads each raster image, and saves converted files into folders that mirror the website page structure. It supports img tags, srcset, picture sources, page icons, and inline CSS url(...) references. When a page provides multiple responsive sizes for the same image, it keeps the highest-resolution candidate instead of downloading every size.

Key Features: - Downloads image references from a website URL and linked same-site pages - Collapses responsive variants to the highest-resolution image URL - Stores homepage images in the output root and subpage images in matching folders - Converts images to WebP, PNG, JPG, or JPEG - Works interactively or with CLI flags for scripts

Quick Start¶

Run Instantly (No Installation)¶

uvx --from git+https://github.com/pruthivithejan/r10n.git r10n website-images

Run Locally¶

git clone https://github.com/pruthivithejan/r10n.git
cd r10n
uv sync
uv run r10n website-images

Usage¶

Interactive Mode¶

Run without arguments for step-by-step prompts:

uv run r10n website-images

Command Line Mode¶

Pass all options directly:

uv run r10n website-images \
  --url https://example.com \
  --output local/outputs/website-images \
  --format webp \
  --quality 85 \
  --max-pages 50 \
  --yes

Parameters¶

Examples¶

Example 1: Basic Usage¶

uvx --from git+https://github.com/pruthivithejan/r10n.git r10n website-images \
  --url https://example.com \
  --max-pages 50 \
  --yes

Example 2: Convert to PNG¶

uvx --from git+https://github.com/pruthivithejan/r10n.git r10n website-images \
  --url https://example.com \
  --format png \
  --output downloads/example-images \
  --max-pages 25 \
  --yes

Example 3: Local Installation¶

uv run r10n website-images \
  --url https://example.com \
  --output local/outputs/website-images/example \
  --format jpg \
  --quality 90 \
  --max-pages 0 \
  --yes

Input Format¶

The input is a website URL. The command scans that page and follows same-site links found in <a href="../..."> elements:

https://example.com

The command scans the HTML for image references in:

<img src="/image.png">
<img srcset="/small.jpg 1x, /large.jpg 2x">
<source srcset="/photo.webp 800w">
<link rel="icon" href="/favicon.png">
<div style="background-image: url('/background.jpg')"></div>
<a href="/about">About</a>
<a href="/tours/rafting/">Rafting</a>

Output¶

• Output location: local/outputs/website-images/
• Output format: converted image files in the selected format
• Homepage images: saved directly in the output directory
• Subpage images: saved in folders matching the page path
• File naming: numbered files based on the source filename, such as 001-hero.webp

Example output:

local/outputs/website-images/
├── 001-home-hero.webp
├── 002-logo.webp
├── about/
│   └── 001-team.webp
└── tours/
    └── rafting/
        ├── 001-rafting-hero.webp
        └── 002-river.webp

Responsive variants such as ?w=256, ?w=2048, or filename sizes like photo-320x180.jpg and photo-1920x1080.jpg are grouped so only the largest version is downloaded. Unsupported image downloads, such as SVG files or broken URLs, are reported as failed while the rest of the images continue processing.

Troubleshooting¶

Common Issues¶

Issue: Some images are missing - Cause: The website may load images with JavaScript after the page loads. - Solution: Use the static page URL that contains the image references, or download those assets separately.

Issue: Not all website pages were scanned - Cause: The --max-pages limit was reached, or some pages are not linked in static HTML. - Solution: Increase --max-pages or use --max-pages 0 for no limit.

Issue: Unsupported image format - Cause: The source image may be SVG or another format Pillow cannot convert. - Solution: Use source raster images such as PNG, JPG, JPEG, or WebP.

Issue: Website could not be fetched - Cause: The URL is invalid, the site is down, or the request timed out. - Solution: Include https:// in the URL and try increasing --timeout.

See Also¶

• Images
• Get Started Guide