Home › Blog › X to Markdown 2026: 3 Methods

X to Markdown 2026: 3 Ways to Save X Threads & Articles

2026-06-08 · 8 min read · by ThreadGrab

X (formerly Twitter) hosts tens of millions of public threads, articles, and discussions every day. If you are a researcher, writer, or AI practitioner who wants to save that content for offline reading, analysis, or LLM training data, you need a reliable way to convert X content to clean Markdown.

The good news is that 2026 offers at least three solid options, each suited to a different workflow. This guide covers every method in detail so you can pick the one that fits your needs -- whether you are grabbing a single hot take for ChatGPT or running a nightly script that archives every thread from your favorite authors.

What Changed in 2026 for X Content Archiving?

Two things made X-to-Markdown significantly easier this year. First, tweet.md launched a dead-simple service: replace x.com with tweet.md in any post URL and get clean Markdown back, optimized for LLMs. It works with individual posts, threads, and even profile pages. Second, the wider LLM ecosystem -- ChatGPT, Claude, Gemini CLI, Hermes Agent, Cursor -- has made Markdown the default interchange format for training data and context windows, driving demand for clean, token-efficient text extraction.

Method 1: tweet.md -- The One-Liner Conversion

tweet.md is the simplest way to convert an X post to Markdown. You do not install anything -- just edit the URL. Every post and profile on X has a public URL; changing the domain from x.com to tweet.md gives you a Markdown-rendered version of the same content.

# Original X post URL
#   https://x.com/paulg/status/1234567890123456789
#
# Just replace x.com with tweet.md:
#   https://tweet.md/paulg/status/1234567890123456789
#
# The page loads as clean Markdown. Copy it and paste
# into ChatGPT, Claude, or any Markdown editor.

The output includes the author name, post text, engagement stats, and in the case of threads, the entire thread is concatenated as a single Markdown document. tweet.md also offers a skill file for AI agents (install via npx skills add tweet-md/skill) so that Claude Code, Cline, or Hermes Agent can fetch X content on the fly.

tweet.md was recently featured on Product Hunt. It is especially useful when you are in an LLM chat and come across an X thread you want the model to analyze -- replace the URL, copy the Markdown, and paste it in.

When to use tweet.md

• You read X on your phone or laptop and want to save a single thread for later.
• You want to feed an X thread into ChatGPT or Claude for summarization or analysis.
• You want to share a clean, formatted version of an X thread with a colleague without screenshots.

Method 2: ThreadGrab Public API -- Programmatic Bulk Archiving

ThreadGrab exposes a public API that returns X thread content as structured JSON. This is the right choice when you need to automate: batch-fetch threads from your favorite authors, archive them to disk every night, or build a custom research pipeline.

# Fetch a user's recent threads as JSON
curl -s https://threadgrab.com/api/profile/paulg \
  | jq '.[:3] | .[] | {author, text_length: (.text | length)}'

# Save the full thread as Markdown to a local file
curl -s https://threadgrab.com/api/profile/paulg \
  | jq -r '.[] | "# \(.author)\n\n\(.text)\n---"' > paulg-archive.md

ThreadGrab's API returns thread data in a predictable schema: author handle, display name, post text, engagement stats, and timestamps. You can process this with any scripting language -- Python, Bash, Node.js -- and convert it to Markdown, JSON, or CSV.

Unlike tweet.md, ThreadGrab also supports Bluesky posts via the AT Protocol's open API. If you archive content from multiple social platforms, ThreadGrab is your single entry point.

When to use the ThreadGrab API

• You maintain a research archive of X discussions and want nightly batch updates.
• You are building a dataset for LLM training or fine-tuning.
• You want to archive both X and Bluesky content through one interface.
• You need structured data (timestamps, engagement counts) in addition to text.

Method 3: Browser Reader Mode and Read-Later Services

This is the lowest-effort method. Every major browser has a reader mode (Ctrl+Shift+R on desktop, or the page icon in the address bar on mobile). When you open an X Article or a thread's HTML view, reader mode strips the sidebar, suggested posts, and ads, leaving just the content.

# Keyboard shortcuts for reader mode:
#   Chrome / Edge:   F9 or Ctrl+Shift+R
#   Firefox:         F9
#   Safari:          Cmd+Shift+R (Reader button in toolbar)
#
# Once in reader mode, copy-paste the clean text into
# your favorite Markdown editor or note-taking app.

Dedicated read-later services like Pocket and Matter go a step further: they save the article to your library, sync across devices, and often expose a highlight/annotation layer. Both Pocket and Matter let you export saved articles as plain text or HTML, which you can then convert to Markdown with a tool like pandoc.

When to use reader mode / read-later services

• You only save X content occasionally and do not want to learn a new tool.
• You already use Pocket or Matter for blog posts and want a unified reading list.
• You are reading X Articles (long-form) rather than threaded short posts -- reader mode handles long-form much better than short-thread concatenation.

Side-by-Side Comparison

Building an Automated Workflow with All Three

Many power users combine methods for a complete pipeline. Here is a real-world example: a researcher who follows 20 X accounts and wants an LLM-ready archive delivered to their Obsidian vault every morning.

#!/bin/bash
# Nightly X archive for Obsidian (runs at 6 AM via cron)

USERS=("paulg" "kelseyhightower" "swyx" "levelsio")
OUTPUT_DIR="$HOME/obsidian-vault/inbox/x-threads"
mkdir -p "$OUTPUT_DIR"

for user in "${USERS[@]}"; do
  curl -s "https://threadgrab.com/api/profile/$user" \
    | jq -r '.[] | "# \(.author)\n\(.text)\n---"' \
    > "$OUTPUT_DIR/$user-$(date +%Y-%m-%d).md"
done

echo "Archived ${#USERS[@]} profiles to $OUTPUT_DIR"

This pipeline uses ThreadGrab's API for batch-fetching, outputs clean Markdown that Obsidian can index, and leaves the door open for tweet.md when the researcher spots a single interesting thread during the day and wants it in their LLM chat immediately.

How ThreadGrab Fits Into These Workflows

ThreadGrab was built to give you ownership of the content you read. Whether you use it through the public API (for automation), the web interface (for one-off grabs), or as a complement to tweet.md (for LLM workflows), the goal is the same: convert public social content to clean, portable, machine-readable Markdown that you own and control.

Unlike tweet.md, ThreadGrab does not require an account, an API key, or any payment for programmatic access. It also works with Bluesky posts, making it the only tool in this comparison that spans two major social platforms.

FAQ

Pick Your Method and Start Saving

The X-to-Markdown ecosystem in 2026 is healthier than ever. tweet.md gives you a one-second conversion for any post. ThreadGrab's public API gives you free, scriptable access for research and automation. Browser reader mode and read-later services cover the casual save. Choose the method that matches your workflow, and if you find yourself wishing for a simpler way to save threads and articles for later reading, ThreadGrab has you covered.