50 MB Avro sample

An Avro file of exactly 52,428,800 bytes. It is a valid Apache Avro object container file file, generated by this site. Compare its checksum and test it with the parser used by your application. Use it wherever you need a 50 MB sample, test or dummy Avro file.

sample-avro-50mb.avro

50 MB52,428,800 bytes

Download 50 MB Avro

Created in your browser and, in browsers that can stream to disk, sent straight to your downloads list; this can take a minute. No sign-up.

Type: application/avro
SHA-256: 42c7d895e1b3f0f8fb2937c1a5b00fea27c8fc692d7808c624469dab7c475a12
Delivery: Created in your browser from static generator assets (how it works).

What is in this file

An object container file whose header carries the writer schema as JSON, followed by 419,201 binary-encoded customer records in 4,193 blocks. The last record's optional last_login string is lengthened to bring the file to the exact size.

Records	419,201
Blocks	4,193, of up to 100 records
Schema	Customer record, 12 fields
Codec	null (not compressed)
Sync marker	16 bytes after every block

First bytes of the file

The first 64 bytes of sample-avro-50mb.avro as a hex dump: offset, bytes, then the same bytes as text. The Avro signature is 4F 62 6A 01.

00000000  4f 62 6a 01 04 16 61 76  72 6f 2e 73 63 68 65 6d |Obj...avro.schem|
00000010  61 d0 0a 7b 22 74 79 70  65 22 3a 22 72 65 63 6f |a..{"type":"reco|
00000020  72 64 22 2c 22 6e 61 6d  65 22 3a 22 43 75 73 74 |rd","name":"Cust|
00000030  6f 6d 65 72 22 2c 22 6e  61 6d 65 73 70 61 63 65 |omer","namespace|

Preview

Avro files cannot be shown in a browser. Download the file and open it in an application that reads Avro, such as Apache Spark, Flink and Hive.

Where 50 MB matters

A 52,428,800-byte file: exactly the size at which Git hosting on GitHub starts warning about large files, and about the limit for a zipped AWS Lambda package uploaded directly.

All 50 MB sample files

Use this file for

Measuring download speed from your own server or CDN
Load-testing readers with many blocks
Testing ingestion jobs and bulk loaders

Testing notes for Avro

No media type is registered for Avro, so files arrive labelled application/octet-stream as often as anything else. Detection has to rely on the four signature bytes.
A logical type is an annotation on a plain type: a date is stored as an int counting days since 1970. A reader that ignores the annotation returns a number where a date was expected.

Avro format reference

Technical details

File name	sample-avro-50mb.avro
Format	Avro (Apache Avro object container file)
Extension	.avro
MIME type	application/avro
Size	52,428,800 bytes (50 MB)
Kind	Binary
Signature	4F 62 6A 01
SHA-256	42c7d895e1b3f0f8fb2937c1a5b00fea27c8fc692d7808c624469dab7c475a12
Created	2026-10-23
Validation	See the listed size and SHA-256. Small downloads are checked in the browser; verify streamed downloads manually. A hash match does not prove parser compatibility.
Preview	Not available for this file; details only
Source	Generated in your browser by this site's generator; not copied from elsewhere
Licence	Free to use, modify and redistribute (terms)
Category	Data

Check the stated structure and checksum against your application requirements. As with any download, scan the file with your own security tools before opening it. How files are made and checked

Verify your copy

This file is created in your browser, so it has no direct URL for curl or wget. The generator is designed to produce reproducible bytes. After saving, compare the checksum with the SHA-256 value above.

sha256sum sample-avro-50mb.avro     # Linux
shasum -a 256 sample-avro-50mb.avro # macOS

On Windows, run Get-FileHash sample-avro-50mb.avro -Algorithm SHA256 in PowerShell. For scripts and pipelines, save the file once and keep your own copy.

Related files

Questions about this file

Is this file exactly 50 MB?: Yes. It is 52,428,800 bytes, which is 50 MB in the binary units that operating systems and server limits use (1 KB = 1,024 bytes).
What is inside the 50 MB Avro file?: An object container file whose header carries the writer schema as JSON, followed by 419,201 binary-encoded customer records in 4,193 blocks. The last record's optional last_login string is lengthened to bring the file to the exact size.
Are the Avro samples compressed?: No. They use the "null" codec, which every Avro library is required to support, so the records can also be inspected in a hex viewer.

Download