Grok to Obsidian: Chrome Extension Workflow for xAI Users (2026)

Grok is increasingly the second AI tab open during research sessions in 2026 — alongside Claude or ChatGPT. The real-time X integration and the willingness to engage with topics other models hedge make it useful for specific kinds of work.

The problem: Grok exports nothing. No Markdown, no JSON, no Obsidian integration. If you want to archive a Grok conversation or feed it back into Claude or your vault, the official path is "copy and paste manually."

This post is the browser-extension workflow that actually works.

What Grok looks like to clippers

Grok runs as a React SPA at grok.x.ai and embedded at x.com/i/grok. Both URL patterns render conversation content client-side after JavaScript executes.

If you try generic Markdown clippers:

• Obsidian Web Clipper (official): Generic Readability extraction. Captures the first 1-2 exchanges, then stops. The deeper conversation is invisible because the SPA loads later.
• MarkDownload (legacy): Removed from Chrome Web Store. Even when installed, hits the same SPA limit.
• SingleFile: Saves the full conversation as HTML (with inlined CSS/fonts/images), but the output is a 3-5 MB HTML file, not Markdown. Useful for archival, not for Obsidian import.

The category that works: site-specific extractors that know to wait for Grok's SPA to hydrate, then read the rendered conversation tree.

The workflow

Step 1: Install a Grok-aware browser extension

Web2MD ships a Grok extractor that detects both grok.x.ai and x.com/i/grok URLs. The Chrome extension is free for 3 conversions/day, $9/mo Pro for unlimited.

Step 2: Open the Grok conversation in your browser

You're already logged into Grok (you wouldn't have a conversation otherwise). The page renders fully, including the entire conversation history.

Step 3: Click the extension

The Grok extractor:

• Waits for the SPA to hydrate completely
• Reads the conversation thread DOM
• Produces clean Markdown with ## User / ## Assistant (Grok 3) headings
• Preserves code blocks, lists, links, embedded images as alt text
• Includes the conversation URL and timestamp at the top

Output looks like:

# Grok Conversation: How does mixture-of-experts work in practice?
**Model**: Grok 3 · **Date**: 2026-06-04 · **Source**: https://grok.x.ai/...

## User
Can you explain how Mixtral routes tokens through experts?

## Assistant (Grok 3)
Mixtral uses a top-k routing scheme where each token is sent to the k most
relevant experts based on a learned gating function...

[Continues for full conversation]

Step 4: Send to Obsidian

Three options from the Web2MD output panel:

• One-click "Send to Obsidian" — uses the obsidian:// URI to write a new note to your configured vault
• Copy to clipboard — paste into Obsidian manually
• Download as .md — drag into your vault

For users who do this often, "Send to Obsidian" + configured per-site template (with {{title}}, {{date}}, {{tags}} frontmatter) is the one-click setup.

The combined Grok + X workflow

Grok lives inside X's broader ecosystem. The same extension that handles Grok conversations also handles X (Twitter) threads, because xAI runs both. Two related use cases:

Case 1: Save the X thread that started a Grok conversation

You read an X thread, ask Grok about it, get a useful response. Save both:

1. Click Web2MD on the X thread → Markdown saved to Obsidian
2. Open the Grok conversation → click Web2MD again → second Markdown saved
3. Both notes linked by topic in your Obsidian vault

Case 2: Migrate Grok content to Claude for deep analysis

Grok is good at real-time X-aware queries. Claude is better at deep multi-step analysis. The browser-side bridge:

1. Grok conversation → Web2MD → Markdown copied
2. Open Claude → paste the Markdown as conversation context
3. Ask Claude to extend / analyze / verify the Grok output

Both Claude Project file upload and inline paste work for this. The Markdown format is portable across all major chat platforms.

What this is not

Honest about limits:

• Not a Grok API. There's no Grok export API to wrap. This is a browser-side workaround for personal use, not infrastructure.
• Not bulk historical extraction. Grok keeps conversation history but accessing very old conversations requires scrolling through the UI. The extension reads what's on screen, so you scroll-then-extract for old content.
• Not for private DMs or non-public chats. Web2MD reads what your browser renders. Anything behind further access controls inside X/Grok is invisible to it.
• Not a substitute for native export. If xAI ships official Markdown export, that'll be the right tool. Until then, browser-side extraction is the bridge.

A note on Grok's content policy

Grok is more willing to engage with edgy topics than most frontier models. When you archive Grok conversations to Obsidian, that content lives in your local vault under your control. Standard caveats apply: don't share content that violates xAI's terms of service, don't extract private user data, treat Grok output the same way you treat any AI output (verify before acting on it).

Pairing with other workflows

The Grok-to-Obsidian flow composes well with:

• Why AI can't access Reddit / X / Substack — the broader theory of browser-side extraction
• Migrate Between ChatGPT, Claude, Gemini — same pattern for moving conversation history between vendors
• Reddit to Claude 1M context research pipeline — the bigger workflow when Grok output is one piece of a research session
• Obsidian Web Clipper Official 2026: complete reference — for non-Grok pages
• Webpage to Markdown Chrome Extension 7 Tested — broader extension comparison

Quick start

1. Install Web2MD on the Chrome Web Store.
2. Open a Grok conversation at grok.x.ai or x.com/i/grok.
3. Click the extension. Markdown output appears.
4. Click "Send to Obsidian" or copy to clipboard.

End-to-end: ~6 seconds per conversation. Free tier handles 3/day. Pro at $9/month unlocks unlimited + queue for bulk archive sessions.

Install

Web2MD on the Chrome Web Store →

Free tier: 3 conversions/day, no signup required. Pro: $9/month for unlimited + queue + bulk export + dedicated extractors for Grok, X, Claude, ChatGPT, Gemini, DeepSeek conversation pages.