Save webpages offline with full source code & assets. Snapshot any page with pixel-perfect accuracy or crawl multiple pages into offline-ready ZIP files. Captures everything: images, fonts, SVGs, videos, CSS, animations, forms, tables, and more.
Works on all major browsers.
99%+ Site Compatibility - Landing pages, web apps, documentation, e-commerce, news sites, and complex SPAs.
- Full support via modern extension APIs (Manifest V3)
- Native WebExtension support across all platforms
- Unified codebase works on all major desktop browsers
- @container: Container queries for responsive components
- @layer: CSS Cascade Layers for style organization
- @supports: Feature queries for progressive enhancement
- @scope: CSS Scope for component isolation
- @media print: Print stylesheet preservation
- @scroll-timeline: Scroll-driven animations
- @counter-style: Custom list markers
- @page: Print page settings
- @font-feature-values: OpenType font features
- Plus existing: @keyframes, @property, @font-face
- Recursive shadow root traversal: Captures content from open shadow roots
- Shadow DOM assets: Images, backgrounds, SVGs from web components
- Declarative Shadow DOM: Handles
<template shadowroot>
- All input types: text, email, checkbox, radio, date, range, color, etc.
- Textarea content: Preserved as textContent
- Select elements: Selected option marked
- Contenteditable: innerHTML preserved
- Datalist options: Autocomplete suggestions kept
<dialog>: Open/closed state preserved<details>/<summary>: Accordion states (FAQs, collapsibles)<progress>/<meter>: Value preservation<output>: Calculated form values<map>/<area>: Image map coordinates<template>: Content detection- Popover API:
[popover]element states - Inert attribute: Non-interactive region marking
- YouTube embeds: Thumbnail placeholders with video link
- Vimeo embeds: Placeholder with link
- PDF embeds: Download link fallback
- Generic iframes: Styled placeholders with hostname
- Clip-path URLs: External SVG clip-path references
- Mask images: CSS mask-image external URLs
- Noscript content: Made visible for offline viewing
- Color scheme: Theme-color and prefers-color-scheme preserved
- ARIA states: expanded, selected, checked, hidden
- Inert regions: Preserved for screen readers
- Semantic HTML: All HTML5 elements maintained
- Skips tracking scripts: Google Analytics, GTM, Facebook Pixel, Hotjar, etc.
- Skips analytics: Mixpanel, Segment, Amplitude, FullStory, etc.
- Skips third-party widgets: Intercom, Drift, Zendesk, Crisp, etc.
- 40+ patterns: Automatically filters scripts that would fail offline
- Faster captures: No wasted time on assets that will be blocked
- Sticky headers: Position and styles preserved
- Virtualized tables: Transform removal for complete capture
- Cell dimensions: Width preservation
- UTF-8 encoding: Charset meta tag injection prevents garbled characters
- 10s timeout per asset (was 30s) - prevents hanging on blocked resources
- 15s hard limit - ensures no single asset blocks the entire capture
- Early abort: Stops retrying when resources are blocked by ad blockers
- Progress tracking: Shows skipped assets count during download
- Comprehensive asset discovery: srcset, picture elements, Open Graph images, Twitter cards
- All CSS properties: mask-image, clip-path, cursor, filter, border-image, shape-outside
- Pseudo-element capture: ::before and ::after content URLs
- Data attribute scanning: Automatically finds images in data-src, data-bg, data-image, etc.
- Modern image formats: AVIF, WebP, JPEG XL support
- 50,000 max assets (was 2,000)
- 25 concurrent downloads (was 15)
- 500KB embed threshold (was 100KB) - more assets embedded inline
- 2GB max ZIP size (was 1GB)
- 2 optimized fetch strategies (was 4) - faster failure detection
- Background fallback: Uses extension context for CORS-blocked resources
- Cache-first fetching: Uses browser cache when available
- Better error recovery: Continues even when some assets fail
- Video tracks: Captures subtitles and caption files
- Audio elements: Full audio file capture
- All font variants: Every weight, style, and format
- Icon fonts: FontAwesome, Material Icons fully captured
- Crawl entire websites with the new "Crawl site" mode
- Same-domain only crawling prevents unbounded crawls
- Configurable max pages limit (default: 10, up to 500)
- Single ZIP output with all pages and deduplicated assets
- Real-time progress tracking in popup
- CSS Hover States:
:hover,:focus,:activerules extracted and preserved - JS Animation Libraries: Detection for GSAP, Anime.js, Framer Motion, Lottie, ScrollMagic
- Scroll-Triggered Animations: Programmatic scrolling to capture all animation states
- Multi-Frame Canvas: Captures 5 frames from animated canvas elements
- Video Poster Extraction: Automatic poster image capture for video elements
- Extracts styles from styled-components, Emotion, Linaria, JSS
- Preserves dynamic styles that would otherwise be lost
- 25 concurrent downloads for fast asset fetching
- SHA-256 deduplication prevents redundant asset downloads
- URL normalization strips tracking params (utm_*, fbclid, gclid)
- 50,000 max assets for larger sites
- Download or clone this repository
- Open your browser's extension management page and enable Developer mode
- Click "Load unpacked" (or equivalent) and select the
SnapSite/folder - Pin SnapSite in the toolbar for easy access
- Open a normal, finite page (blog, docs, landing page)
- Click the SnapSite icon and select "Single Page"
- Click Start capture and save the ZIP
- Unzip and open
index.html- works completely offline with no console errors
- Open any page on the site you want to capture
- Click the SnapSite icon and select "Crawl Site"
- Set the max pages limit (default: 10)
- Click Start capture to begin crawling
- Progress shows "X/Y pages" as the crawl proceeds
- Final ZIP contains all captured pages with shared assets
- Scope: Same domain only - external links are skipped
- Progress: Real-time page count and progress bar
- Stop anytime: Click Stop to end the crawl early
- Memory warnings: Alerts at 80% memory usage
- Continues in background: Crawl continues even if you close the popup
index.html: DOM snapshot with local asset paths and preserved animationsassets/: downloaded CSS/JS/images/fonts/media/videosreport/README.mdandreport/fetch-report.json: human + machine summaryquick-check.html: loadsindex.htmlin an iframe and summarizes the reportreport/asset-manifest.json: original URL + local path, bytes, mime, sha256
site-capture.zip
├── index.html (first/home page)
├── page-1.html (subsequent pages)
├── page-2.html
├── ...
├── assets/ (deduplicated across all pages)
│ ├── image-abc123.jpg
│ ├── font-def456.woff2
│ └── ...
└── report/
└── README.md
SnapSite creates truly offline-ready captures with zero console errors:
The following script types are automatically skipped during capture:
- Analytics: Google Analytics, GTM, Mixpanel, Segment, Amplitude
- Tracking: Facebook Pixel, Hotjar, FullStory, Mouseflow, Crazy Egg
- Widgets: Intercom, Drift, Zendesk, Crisp, Tawk.to, LiveChat
- Auth: Google Sign-In, Facebook Login, reCAPTCHA
- Error tracking: Sentry, Bugsnag, Rollbar, LogRocket
- A/B testing: Optimizely, VWO
- CSS Inlined: All stylesheets embedded directly in HTML
- Fonts Embedded: Web fonts converted to base64 (up to 500KB each)
- Images Embedded: Assets under 500KB embedded as base64 data URIs
- Large Assets: Saved to
assets/folder with correct references - srcset Preserved: All responsive image variants captured
- UTF-8 Encoding: Charset declaration ensures correct character display
| Asset Type | Captured |
|---|---|
| Images (jpg, png, gif, webp, avif) | Yes |
| SVG (inline, external, sprites) | Yes |
| Fonts (woff, woff2, ttf, otf, eot) | Yes |
| Videos (mp4, webm) + posters | Yes |
| Audio (mp3, wav, ogg) | Yes |
| CSS (external, inline, computed) | Yes |
| Favicons (all sizes) | Yes |
| Open Graph / Twitter images | Yes |
| Background images | Yes |
| Mask images / clip-paths | Yes |
| Cursor images | Yes |
| Subtitles / captions (vtt, srt) | Yes |
| Shadow DOM content | Yes |
| Form states | Yes |
When you open a captured page from file://, browsers enforce strict security:
- CORS blocks loading external scripts/fonts
- CSP policies block inline styles
- Relative paths resolve incorrectly
SnapSite handles all of this automatically so your captured pages just work.
- CSS Keyframe Animations: All
@keyframesrules are extracted and preserved - CSS @property: Modern CSS custom properties with animation support captured
- CSS Container Queries:
@containerrules for responsive components - CSS Cascade Layers:
@layerdeclarations preserved - CSS Scope:
@scoperules for component isolation - Hover/Focus/Active States: Pseudo-class rules captured as
.gi-hover-*classes - Computed Animation States: Animation properties (duration, timing, delay) preserved
- Backdrop Blur & Effects: Modern CSS filter effects like backdrop-filter maintained
- Tailwind Animations: Dynamic animation classes from Tailwind CSS preserved
- Transform & Transition: All transform and transition properties captured
- SVG Animations: Inline SVG animations preserved
- Canvas Elements: Animated canvas elements capture multiple frames
- Video Support: Video elements included with poster images and controls
- Scroll Animations: @scroll-timeline and animation-timeline support
SnapSite detects and notes the presence of:
- GSAP / TweenMax / TweenLite: Animation library states captured
- Anime.js: Animation states preserved
- Framer Motion: React animation data captured
- Lottie / Bodymovin: Animation player detection
- ScrollMagic / ScrollTrigger: Scroll animation states
- AOS (Animate On Scroll): Intersection-based animations
- Automatically detects and expands all carousel slides
- Supported libraries: Slick, Swiper, Bootstrap Carousel, Splide, Keen Slider, Owl Carousel, Flickity, Glide
- All slides displayed vertically in the captured page
- Navigation controls automatically hidden since all content is visible
- Heuristic auto-scrolls until bottom is stable
- Caps: runtime, max assets, ZIP size, and concurrency (configurable in Options)
- Optional denylist in settings to block specific URL patterns
- Options page lets you adjust caps, redact behavior, denylist, and whether to save without prompt
- Theme: Light, Dark, or Auto (system)
- Strip Scripts: Optionally remove scripts and inline handlers for offline safety
- Redaction: Off by default. Enable "Redact authenticated text" to replace sensitive text in saved HTML
- Crawl settings: Max pages, delay between captures, memory warnings
- Animation settings: Hover states, scroll animations, canvas frames
- The extension requests
tabspermission for crawl navigation host_permissions: <all_urls>enables cross-origin asset fetching- Core actions (injecting on the active tab, downloads, storage) work with standard permissions
Works on all major desktop browsers with modern extension API support (Manifest V3).
- Third-party iframes show placeholders with links to original content
- YouTube/Vimeo embeds show video thumbnails with links
- Tracking/analytics scripts are automatically skipped (won't cause errors)
- The popup shows live progress during capture
- For crawls, progress shows "X/Y pages" with real-time updates
- In-page overlay shows status with a Stop button during capture