Save Webpage as Markdown: Chrome Extension Guide (2026)
Save Webpage as Markdown: Chrome Extension Guide (2026)
You want to save the webpage you're reading as a Markdown file. Maybe for your Obsidian vault, maybe to paste into Claude, maybe for your personal archive. The category of Chrome extensions that does this exploded over the last 18 months, and the right pick depends on what comes after the save.
This post is the honest comparison after testing 6 extensions on the same content.
The 6 extensions tested
| Extension | Output type | Primary use case | |---|---|---| | Web2MD | Markdown to clipboard / file / AI / Obsidian | AI workflow handoff | | Obsidian Web Clipper (official) | Markdown direct to vault | Obsidian-first workflows | | MarkDownload | Markdown to file / clipboard | Generic clipper (removed from CWS) | | MarkSnip | Markdown to clipboard | Minimalist clipper | | SingleFile | Single .html file with inlined assets | Pixel-perfect archive (not Markdown) | | Save MD | Markdown to Downloads folder | Simple file save |
How they actually handle "save webpage as Markdown"
The interesting differences emerge at three points: extraction quality, output destination, and post-save workflow.
Web2MD
Behavior: Click the extension on any webpage. A panel appears showing the extracted Markdown plus token counts (GPT-4 and Claude). Buttons for:
- Copy to clipboard (default action)
- Download as .md file
- Send to AI: opens ChatGPT / Claude / Gemini / DeepSeek / Perplexity in a new tab with the Markdown pre-filled
- Send to Obsidian: writes the Markdown to your configured vault via the
obsidian://URI scheme - Add to queue: stack across multiple pages for bulk export (Pro)
Strength: 20+ site-specific extractors (Reddit JSON, X auth-session, YouTube transcript anchors, Wikipedia infoboxes, paywalled Substack, Xiaohongshu, WeChat, Notion). Generic pages also work via fallback extraction.
Limit: Free tier is 3 conversions/day; Pro at $9/mo unlocks unlimited + queue + bulk export + API.
Obsidian Web Clipper (official)
Behavior: Click the extension, configure your default vault on first use, save. Subsequent clicks write directly to your vault without prompting (configurable). Templates with frontmatter variables ({{title}}, {{author}}, {{date}}, {{url}}, {{content}}) supported.
Strength: Direct vault writes are the one-click workflow Obsidian users want. Per-site rules let you assign different templates to arxiv.org vs personal blogs.
Limit: Generic Readability extraction. Reddit threads stop at the first "continue this thread" link. YouTube transcript template breaks on every YouTube UI update. No AI-handoff workflow. Known silent-data-loss bug if you cancel the "Open in Obsidian?" dialog (GitHub issue #828, open since 2024).
MarkDownload (legacy)
Behavior: Click the extension. Markdown output appears in a popup with options to copy, download as .md, or save to a specified folder. Custom Turndown rules supported.
Strength: When it worked, simple and effective for public pages.
Limit: Removed from Chrome Web Store in 2025. New installs not possible. GitHub repo abandoned 2 years. Known bugs (residual span tags in code blocks, broken tables, dropped LaTeX, Brave incompatibility) won't be fixed. If you're currently using it, plan migration to Obsidian Web Clipper or Web2MD.
MarkSnip
Behavior: Click the extension, Markdown copies to clipboard automatically. Minimalist UI.
Strength: Lowest-friction setup. Good for users who just want "page → clipboard" without configuration.
Limit: Generic extraction only. Reddit Shadow DOM problems, X SPA timing issues, YouTube transcript breakage — all the standard generic-extraction failures. No AI integration, no queue, no file save.
SingleFile
Behavior: Click the extension, saves the entire webpage (with inlined CSS, fonts, base64-encoded images) as a single .html file in Downloads.
Strength: Pixel-perfect archive. The saved page looks exactly like the original when opened months or years later. Open-source, no telemetry.
Limit: Output is HTML, not Markdown. For AI input you'd need to convert HTML to Markdown as a second step. For pure archival or compliance use cases, it's the right tool. For "save then paste into Claude," not the right shape.
Save MD
Behavior: Click, .md file saves to Downloads. No popup, no preview, no further options.
Strength: Truly minimal. Good for users who want a one-click "save as file" workflow.
Limit: Generic extraction. No queue, no AI integration, no Obsidian URI. The file lands in Downloads and you move it manually.
Recommendation by destination
Destination: Obsidian vault
Use Obsidian Web Clipper (official). It's free, open-source, maintained by the Obsidian team, and direct vault writes are the one-click flow you want. The known limitations are real but manageable for most Obsidian-first workflows.
For Reddit / X / Substack / Xiaohongshu where Obsidian Web Clipper's generic extractor produces partial output, add Web2MD alongside.
Destination: Claude / ChatGPT / Gemini / Cursor
Use Web2MD. The send-to-AI button shortcuts the copy/paste round trip. Built-in token estimation lets you budget context before sending. Site-specific extractors handle the platforms generic clippers fail on.
Destination: Local file archive
Use Save MD or Web2MD's download option. Both produce .md files in Downloads. Web2MD also has bulk export for combining many pages into one file.
Destination: Pixel-perfect HTML preservation
Use SingleFile. Different category (HTML not Markdown) but the right tool when visual fidelity is what you actually want. Common scenario: legal evidence preservation, research provenance, "I want to see this page exactly as it appeared on this date."
Destination: Quick clipboard copy
Use MarkSnip or Web2MD's default copy action. MarkSnip is the simpler setup; Web2MD adds the option ladder.
The 2-extension stack most heavy users run
After interviewing ~30 heavy clipper users, the typical setup is:
- Obsidian Web Clipper (official) for vault writes
- Web2MD for AI workflow handoff and platform-specific extractors
The two cover complementary destinations. Same toolbar, different button per destination.
For users with simpler needs (just save a page now and then), pick one based on the primary destination above.
What about Edge / Firefox / Brave?
- Edge: Chrome extensions install directly from CWS via Edge's "Allow extensions from other stores"
- Firefox: Most extensions ship a Firefox version separately. Web2MD, Obsidian Web Clipper, SingleFile all available
- Brave: Native Chrome extension support. Web2MD ships native Brave compatibility; Obsidian Web Clipper works with some Brave Shields adjustments
Pairing with other workflows
The save-webpage workflow composes with:
- Webpage to Markdown Chrome Extension: 7 Tested — deeper comparison with more extensions
- Why AI can't access Reddit, X, Substack — why browser-side matters
- Best Web Clipper for Obsidian + AI in 2026 — Obsidian-specific picks
- Obsidian Web Clipper Official 2026: Complete Reference — deep dive on the official tool
- Save webpage as Markdown — supported sites page — all 35+ sites Web2MD has dedicated extractors for
Quick start
- Pick the destination first (vault, AI chat, file, archive).
- Install one extension matching that destination from the comparison above.
- Click on any webpage to save.
Total time from "I want to save this page" to "saved in my workflow": ~5 seconds per page once configured.
Install
Web2MD on the Chrome Web Store →
Free tier: 3 conversions/day. Pro: $9/month for unlimited + queue + bulk export + dedicated extractors for 20+ platforms.