Obsidian Web Clipper Official 2026: Complete Reference + Companion Tools
Obsidian Web Clipper Official 2026: Complete Reference + Companion Tools
The official Obsidian Web Clipper shipped in late 2024 and is now the default save-to-vault tool for Obsidian users in 2026. If you searched for "obsidian web clipper official 2026" looking for the docs, install link, template syntax, or a status update on known issues — this post is the reference.
It is also honest about what the official clipper does not do, and what to add alongside it for AI workflows.
Official sources
| What | URL | |---|---| | Official landing page | https://obsidian.md/clipper | | Chrome Web Store | https://chromewebstore.google.com/detail/obsidian-web-clipper/cnjifjpddelmedmihgijeibhnjfabmlf | | Firefox Add-ons | https://addons.mozilla.org/en-US/firefox/addon/web-clipper-obsidian/ | | Safari (App Store) | https://apps.apple.com/us/app/obsidian-web-clipper/id6720708363 | | Microsoft Edge Add-ons | https://microsoftedge.microsoft.com/addons/detail/obsidian-web-clipper/eigdjhmgnaaeaonimdklocfekkaanfme | | GitHub repo | https://github.com/obsidianmd/obsidian-clipper | | Issues & feature requests | https://github.com/obsidianmd/obsidian-clipper/issues | | Discord support | Obsidian Discord, #plugin-clipper channel |
There is no separate help center or wiki — the GitHub README plus open issues are the canonical docs in 2026.
What the official Obsidian Web Clipper does well
After 6 months of daily use, the strengths I can vouch for:
Direct vault writes
One click puts a .md file in your Obsidian vault, in the folder you configured, with the template you defined. No manual copy-paste cycle. For "clip into second brain" workflows this is the entire workflow.
Template system
You can define per-page-type templates with frontmatter variables:
---
source: {{url}}
author: {{author}}
published: {{published}}
clipped: {{date}}
tags: [webclip, {{domain}}]
---
# {{title}}
{{content}}
Per-site rules let you assign different templates to arxiv.org vs personal blogs vs Wikipedia.
Cross-browser
Chrome, Firefox, Safari, Edge — all maintained by the official team. No "this works on Chrome but breaks on Firefox" issues that smaller clippers struggle with.
Reader-mode extraction
Strips the boilerplate (nav, ads, footer, related-article widgets) using Readability-style heuristics. Output is clean for most blog and news content.
Privacy-first
Runs locally in your browser. No content sent to third-party servers. The vault write happens via obsidian:// URI scheme to your local Obsidian app.
What the official clipper struggles with (mid-2026)
Documented limits from the GitHub issues tracker and my own testing:
1. Silent data loss on dialog cancel (GitHub #828, open)
When you cancel the "Open in Obsidian?" dialog the clip is silently discarded. No notification, no clipboard fallback, no draft saved. This is the bug that drives users to companion tools — the cancel button is one tap away from accidental clicks.
Status as of June 2026: open, no PR merged. Workaround: don't cancel; if you change your mind about destination, edit the file in Obsidian after import.
2. Reddit Shadow DOM problems
Modern Reddit renders comment trees in Shadow DOM. The official clipper extracts the OP post and the first 1-2 visible comments, then stops at the next 'continue this thread' link. For threads with 100+ comments — where the substance lives — most of the discussion is silently lost.
Workaround: use a Reddit-aware extractor like Web2MD that hits Reddit's .json endpoint for the full tree (see Why Claude can't read Reddit).
3. YouTube transcript breakage on UI updates
The YouTube template depends on YouTube's UI DOM, which Google rebuilds every few weeks. After each YouTube redesign, the transcript template returns empty until users patch the selectors and PR them in. Currently functional as of v0.9 (June 2026), but the cycle repeats.
Workaround: use a dedicated YouTube extractor that pulls from the YouTube Transcript API directly (see YouTube transcript to Markdown).
4. 2000-character selection limit (Chrome + Windows)
A Chrome platform quirk — text selections over 2000 characters get truncated when passed to extension context scripts. For clipping a long quote from a research paper, this can silently lose half the selection.
Workaround: use full-page clip instead of selection on Chrome/Windows, then trim in Obsidian.
5. No send-to-AI workflow
The plugin saves to your vault. It has no concept of "send this article to ChatGPT" — you'd open the saved note in Obsidian, copy, paste manually into your AI chat. For users who paste webpages into Claude daily this adds 10-15 seconds per clip.
Workaround: pair with an AI-handoff clipper for that specific moment.
6. No site-specific extractors
The plugin uses one generic Readability pipeline for all sites. Sites with unusual DOM structures (Stack Overflow code blocks, Wikipedia infoboxes, X/Twitter SPA timing) produce sub-par output. Not a blocker for most use cases but adds cleanup time.
Companion tools that fill these gaps
The realistic 2026 setup for serious clippers is 2-3 tools, not one. Here is what most heavy users I know run:
For "clip into Obsidian vault" workflows
Obsidian Web Clipper (official). The right default. Use it for everything that ends up in your vault.
For "clip into AI chat" workflows
Web2MD (web2md.org). Fills the send-to-AI gap with one-click ChatGPT/Claude/Gemini/DeepSeek/Perplexity handoff. Ships 20+ site-specific extractors (Reddit JSON, X auth session, YouTube transcript anchors, Wikipedia infoboxes, Stack Overflow code-block cleanup, Chinese platforms like Xiaohongshu and WeChat). Free tier: 3 conversions/day, no signup.
For "archive the page exactly as it looked"
SingleFile (open source). Saves the complete page as one self-contained HTML file with inlined CSS/fonts/images. Complementary to both clippers — use when the visual fidelity matters more than Markdown structure.
For "reading queue with highlights"
Readwise Reader ($9.99/mo). If your workflow is read + highlight + sync-highlights-to-Obsidian, Readwise is the right add. Doesn't compete with the clippers — runs alongside.
Most serious users I know run Obsidian Web Clipper + Web2MD as the daily pair. Different jobs, no overlap.
Practical setup recipe (Obsidian + AI workflows)
Step-by-step for the combined setup:
- Install Obsidian Web Clipper from your browser's extension store. Configure your default vault and a basic template with frontmatter (title, source, date, tags).
- Install Web2MD from the Chrome Web Store. No account required for the free tier.
- Pin both to the toolbar. Visual hint: official clipper has Obsidian's purple icon, Web2MD has a stylized "MD" mark.
- Decide by destination, not by tool:
- Page is going into your vault → official Obsidian Web Clipper
- Page is going into an AI chat → Web2MD
- Page is going into both → official clipper first (canonical copy in vault), Web2MD second to send-to-AI
- Configure per-site templates in the official clipper for the sources you clip most (arxiv.org, news, Wikipedia, your favorite blogs).
- Use Web2MD's queue feature for research sessions where you collect 20+ URLs to paste as one Markdown corpus into Claude.
Frequently asked questions
Is the official Obsidian Web Clipper free?
Yes. Free, open source, no premium tier. Funded by Obsidian's broader business model (Sync subscriptions, Publish, Commercial Use License). The plugin team accepts community contributions via GitHub.
Does it work on iOS Safari?
The Safari extension covers macOS and iPadOS. iOS Safari has it via the App Store version. Direct vault writes work on devices where Obsidian itself is installed.
Can I sync templates across devices?
Templates are stored per-browser-installation. There's no built-in cloud sync of templates yet. Workaround: export your templates as JSON and import on each device.
How do I report a bug?
GitHub Issues at https://github.com/obsidianmd/obsidian-clipper/issues. The team is responsive but issue backlog grows; for known issues like #828 (silent data loss), workarounds documented above are your best option short-term.
Does the official clipper support a custom Markdown engine?
It uses a forked Readability + custom Markdown serializer. The serializer handles core MD features but not all GFM (e.g., footnotes with caveats). For pages with complex tables, math, or code structure where the default output disappoints, route those through a different tool.
Bottom line
The official Obsidian Web Clipper is the right default for Obsidian-first workflows in 2026. It is not the right tool for AI handoff workflows, or for Reddit/X/Xiaohongshu where the generic extractor produces sub-par output. Pair it with a companion clipper for those, and the combined stack covers what MarkDownload used to do plus the AI workflows MarkDownload never had.
Related
- Why Claude can't read Reddit (and how to fix it)
- Obsidian Web Clipper + Web2MD: complete 2026 stack
- 5 Best Web Clippers for Obsidian + AI in 2026
- Obsidian Web Clipper vs Web2MD comparison
- YouTube transcript to Markdown for Claude
Install
For the official Obsidian Web Clipper: Chrome Web Store →
For the AI-workflow companion: Web2MD on the Chrome Web Store →
Free tier: 3 conversions/day. Pro at $9/mo for unlimited + queue + bulk export + 20+ site-specific extractors.