twitter — offline viewer and archiver for X (Twitter)

A Rust package (library + CLI) for keeping a personal, offline archive of posts and media from X (Twitter): download everything from an account, or individual posts by URL, then browse the archive in a self-contained HTML viewer — no internet required afterwards.

This package is not affiliated with, endorsed by, or associated with Twitter, Inc. or X Corp. It is an independent, unofficial tool that uses publicly available web endpoints for personal archiving. Archive content only if you are entitled to keep it, and respect the platform's terms.

What it does

• Archives accounts: fetch a user's timeline (paginating through the entire history, or limited with --max) and save every post plus all attached photos, videos and GIFs in original quality.
• Archives individual posts: from a URL (https://x.com/user/status/123...) or a bare post id.
• Browses offline: generates a self-contained index.html with a search box, embedded post data and locally stored media. Open the file directly, or serve it over HTTP with twitter view --serve.
• Keeps your archive current: re-running merges new posts into the existing archive; files already present are skipped.
• Works for protected accounts: when provided with your session cookies (auth_token / ct0), posts you are allowed to see are archived. Without cookies, public posts are still archived via X's public syndication endpoint.

Install

cargo install twitter

or build from source:

git clone https://github.com/twitter-rs/twitter.git
cd twitter
cargo build --release

CLI usage

Usage: twitter [OPTIONS] <COMMAND>

Commands:
  download  Download posts and media from accounts (@user) or post URLs
  info      Print post or account info as JSON
  view      Build the offline HTML viewer (and optionally serve it)

Download everything from an account

twitter download @nasa

Download individual posts by URL

twitter download https://x.com/nasa/status/2075726968335618499

Options

twitter download @nasa --max 50 --no-retweets          # newest 50 original posts
twitter download @nasa -o nasa_archive                 # custom output dir
twitter download https://x.com/... -c cookies.txt      # custom cookies file
twitter info https://x.com/nasa/status/123             # show post JSON
twitter view -o nasa_archive --open                    # open the viewer
twitter view -o nasa_archive --serve 8080              # serve over HTTP

Output layout

twitter_data/
├── index.html     # offline viewer (open in any browser)
├── posts.json     # all archived posts as JSON (newest first)
└── media/         # downloaded photos / videos / gifs, one file per item

Media files are named <post_id>_<n>.<ext>; videos keep a matching <post_id>_<n>_poster.jpg thumbnail for the viewer.

Library usage

use twitter::{Target, Twitter};

let tw = Twitter::from_cookies_file("cookies.txt")?;   // optional auth
let user = tw.user("nasa")?;
let posts = tw.timeline(&user, Some(20))?;             // newest 20 posts
for post in &posts {
    tw.download_media(post, std::path::Path::new("twitter_data"))?;
}

Public API: Twitter (fetch posts/timelines/users, download media), Store (merge/save archives), viewer (offline HTML viewer + tiny HTTP server), plus the Tweet, User, Media models.

Cookies

Copy your auth_token and ct0 cookies (and ideally the rest of your session cookies) from your browser into a Netscape-format cookies.txt file, e.g.:

.x.com	TRUE	/	TRUE	1801772568	auth_token	<your token>
.x.com	TRUE	/	TRUE	1801772569	ct0	<your token>

--cookies cookies.txt is the default; pass -c <file> to override. Keep the file private — it grants access to your account.

Notes

• GraphQL query ids used by the web client change from time to time; the tool re-extracts the current ones from the live web bundle automatically when a request fails, so archives keep working without updates.
• Photos are saved in original quality (?name=orig); videos are saved as the highest-bitrate MP4 variant.
• Be polite: requests are paced and retried with backoff on rate limits.

License

MIT