An archive.org download works in one of two ways, depending on what you want. Books, films, audio, software and other uploaded items live in the Internet Archive’s library, and each item page has a “Download options” panel that gives you the files directly, or you can fetch whole items and collections with the free ia command-line tool. Old websites live in the Wayback Machine instead, and there is no download button for them: you download captured pages and files one by one, or with a downloader that rebuilds the site. This guide covers both, with the steps, the commands and the limits you should expect.
Two different archives: items and the Wayback Machine
People say “archive.org” for two collections that behave very differently:
| Library items | Wayback Machine captures | |
|---|---|---|
| What it is | Files uploaded as “items”: books, audio, video, software, datasets | Copies of web pages saved by crawlers over time |
| Typical address | archive.org/details/<identifier> |
web.archive.org/web/<timestamp>/<url> |
| How to download | Download options panel, file list, torrent, or the ia tool |
Original-file URLs, the CDX API, or a Wayback downloader |
| What you get | The files as uploaded, often in several derived formats | Whatever the crawler captured: HTML, images, CSS, scripts, some documents |
If you are trying to recover an old website, skip ahead to the Wayback Machine section. If you want a book, recording or film, start with the item steps.
How to do an archive.org download of a single item
- Open the item page. The URL looks like
https://archive.org/details/<identifier>. The identifier is the unique name of the item; you will need it for command-line downloads. - Find “Download options”. On most item pages it sits in the right-hand column and lists the available formats, such as PDF, EPUB, MP3, MP4 or ZIP, with a link for each.
- Use “Show all” for the full file list. This opens a directory listing of every file in the item, including originals and the formats the archive generated from them. Originals are usually the best quality.
- Choose the torrent for large items. Many items offer a torrent file that downloads everything in the item and can resume if interrupted.
Some items cannot be downloaded at all. Books in the lending library, for example, can be borrowed and read in the browser but do not offer the full file. Other items may be restricted by the uploader or removed. If there is no download link, that is a deliberate restriction, not a bug to work around.
Bulk archive.org download with the ia command-line tool
The Internet Archive publishes an official Python package called internetarchive (documentation), which installs a command-line tool named ia. It is the most reliable way to download many files or whole collections.
Install and configure
pip install internetarchive
ia configureia configure stores your archive.org account details for actions that need them. Downloading public items does not require an account, so you can skip it if you only download.
Download a whole item
ia download <identifier>This creates a folder named after the identifier and saves every file in the item.
Download only certain files
ia download <identifier> --glob="*.mp3"
ia download <identifier> --format="Text PDF"
ia download <identifier> <filename>--glob matches file names, --format matches the archive’s format labels (run ia metadata --formats <identifier> to see which exist), and naming a file downloads just that file.
Download a collection or search results
ia download --search="collection:<collection-name>" --glob="*.pdf"This works through every item that matches the search. Re-running the same command skips files you already have, which makes interrupted jobs easy to resume. Check ia download --help on your installed version for the current options, because they change between releases.
Be patient with big jobs
The Internet Archive is a non-profit with finite bandwidth. Very large or very fast jobs can be throttled. Run one download at a time, avoid starting many parallel copies of the tool, and expect large collections to take hours.
How to download a website from archive.org
Websites are a different case. The Wayback Machine stores captures of individual pages and files, each with its own timestamp. There is no single archive.org download button for a whole site, so you assemble the copy yourself:
- Check what was captured. Search the domain on
web.archive.organd look at the calendar, or list every captured URL with the CDX API. Our Wayback Machine API guide shows the queries. - Pick a snapshot period. Choose dates where the pages you need were captured completely; see how to choose the right snapshot.
- Download the original files. Add
id_after the timestamp in each capture URL to get the stored file without the archive toolbar, for exampleweb.archive.org/web/20190101000000id_/https://example.com/. - Or use a downloader. Open-source tools such as wayback-machine-downloader on GitHub automate the listing and downloading. Our walkthrough on how to use Wayback Machine Downloader covers installation and the options that matter.
- Fix links and test. Downloaded pages often still reference
web.archive.orgor absolute paths. Clean them before you call the copy finished.
The full process, from finding captures to a working offline copy, is in our main guide to downloading a website from the Wayback Machine. If you would rather not run these steps yourself, a Wayback site downloader service returns the pages as a cleaned HTML ZIP.
What an archive.org download will not give you
- No databases or server code. A Wayback capture holds the HTML the server produced, not the PHP, the database, user accounts or orders behind it.
- No pages the crawler never saw. Pages behind logins, search results and rarely linked pages are often missing.
- Not always every asset. Images, fonts and scripts can be missing from a capture even when the page text is there.
- No unrestricted access to restricted items. Lending-library books and items the uploader restricted stay restricted.
Common archive.org download problems
| Problem | What it usually means | What to try |
|---|---|---|
| Download is very slow | Archive load or a large file | Use the torrent, or the ia tool and leave it running |
| No download link on the item | Lending-only or restricted item | Borrow and read in the browser; look for another edition |
| HTTP 429 errors | Too many requests from your IP | Stop, wait, then run one download at a time; see our guide to the Wayback Machine 429 error |
| Downloaded web page shows the archive toolbar | You saved the replay page, not the original file | Use the id_ form of the URL |
| Site copy has broken images and links | Assets not captured, or links still point at the archive | Check other capture dates; rewrite links locally |
When a done-for-you download makes sense
Item downloads are usually quick and need no help. Website downloads are where time goes: listing captures, choosing dates, fetching hundreds of files slowly enough to avoid rate limits, then cleaning the code so the copy works offline or on new hosting. Our archive.org website download service delivers a cleaned HTML ZIP with archive code removed and links fixed, and the pricing page lists the packages. If you are not sure the archive holds enough of your site, ask for a free archive check first.
Frequently asked questions
Is an archive.org download free?
Yes. Public items and Wayback captures can be downloaded without paying the Internet Archive. Some items are restricted or lending-only and cannot be downloaded in full.
Do I need an account to download from archive.org?
Not for public files. An account is needed for borrowing lending-library books and for uploading. The ia tool only asks for credentials for actions that require them.
Can I download an entire archive.org collection at once?
Yes, with the ia tool and a --search="collection:…" query. Large collections take a long time; run them as a single, resumable job.
Why does my archive.org download stop or fail?
Usually network interruptions, very large files or rate limiting. Torrents and the ia tool resume more reliably than a browser download.