The Pregenerated File Archive contains contains a collection of files with true random data, generated by RANDOM.ORG over many years. Each file is 8,388,608 bits (one mebibyte) long, and we publish one new file every day. The file names follow the ISO 8601 standard date notation, i.e., YYYY-MM-DD. Each file is available in two formats: binary and text. Neither type of file is intended for viewing in a web browser; instead you probably want to download a file and feed it to another program. For more details, please see the Binary Files or Text Files pages.
The files are useful for people who need large amounts of true random data, for example for simulation or research purposes. In addition, the files are used as the basis for some of the pregenerated randomizations supported for example by the Integer Generator and the Sequence Generator in advanced mode. Hence, if you use one of these generators with a pregenerated randomizaton from a given day, the numbers you get are in reality created using the files listed here.
The files are quite large, and to help cover our bandwidth costs, we therefore ask that you set up an Pregenerated File Archive Subscription if you want to download the pregenerated files directly from our servers. Alternatively, you can download the files for free and without registering via BitTorrent. For more info, please see Q3.
BitTorrent is a protocol to help share large files over the internet. To use it to download files from the Pregenerated File Archive, you first need to install a BitTorrent client and then click on any of the ‘.torrent’ files from the Binary Files or the Text Files pages.
Please do! We are very grateful for your help with this.
A new file is published every day, shortly after midnight (UTC). New files are made available in binary as well as text formats. At the beginning of each month, we package up all the files from the previous month in an archive and make it available for download in ZIP and TGZ formats, as well as via BitTorrent.
There were some days far back in the past where no file was published due to technical problems with the scripts that generated them. The last time this happened was in October 2007, when the Pregenerated File Archive was still in its infancy.
Yes, many people use the files in the Pregenerated File Archive for many different purposes. Popular examples include simulation and modeling applications, or to test compression or encryption algorithms. This means the data in the files is not private to you.
All the randomness generated by RANDOM.ORG is subjected to statistical tests and the results published on our Statistics Pages. If you want to test the data yourself, we recommend the DieHarder test suite. Note that the DieHarder tests require a lot of data, so you'll probably want to download a whole month of files or more, if you want the files for that purpose.
No, you're free to use them for what you want, as long as you don't charge money for them.
The data in the binary files is formatted as a single long string of bytes, i.e., values in the [0,255] interval. You probably don't want to open these in a web browser, but instead download them and feed them to another program.
The data in the text files is formatted as a single long string of bits (0s and 1s) in ASCII character format. There is no spacing or punctuation in the files, not even a terminating line feed. You probably don't want to open these in a web browser, but instead download them and feed them to another program.
Not at this point, unfortunately! If you have some code that you're willing to share, send us a link and we'll set up a page for it.
If you're looking to convert the true random data to different ranges, e.g., to produce coin flips, dice rolls or integers in configurable intervals, you need a ‘scaling algorithm.’ For an introduction to scaling algorithms, please see this article by Doctor Jacques.
We don't offer a script to download the whole archive, but you can download large amounts of files quite easily by clicking on the monthly files on the Binary Files or the Text Files pages.
You can also write your own script to fetch the files using shell commands like the following:
$ wget --user=username --password=secret --auth-no-challenge --content-on-error "https://archive.random.org/download?file=2020-08-14.bin" $ curl --user username:secret "https://archive.random.org/download?file=2020-08-14.bin" > 2020-08-14.bin
(‘$’ is the Unix shell prompt.)
Yes! We offer a File Generation Service that lets you generate your own files with up to 20,000,000 true random values to your custom specification, e.g., alphanumeric promotional codes for printing or decimal fractions for scientific simulation. The files are generated in several formats, including plain text, CSV and Excel.
We are happy to take additional questions! Just drop us an email.