Sample Avro files

Apache Avro is a row-based data serialization format with the schema embedded right in the file. Widely used with Kafka and Hadoop. Below are sample .avro files with various schemas, types and codecs to download

Sample file: Simple Avro: 5 records, basic types
Simple Avro: 5 records, basic types
349 bytes
↓
Sample file: Users: 1000 records, realistic dataset
Users: 1000 records, realistic dataset
49 Kb
↓
Sample file: All types: primitive, enum, fixed, union, array, map, logical (date/timestamp)
All types: primitive, enum, fixed, union, array, map, logical (date/timestamp)
1.3 Kb
↓
Sample file: Nested data: record inside record, array and map
Nested data: record inside record, array and map
571 bytes
↓
Sample file: Large Avro: 100,000 records (deflate codec)
Large Avro: 100,000 records (deflate codec)
735 Kb
↓
Sample file: With deflate codec: 20,000 records
With deflate codec: 20,000 records
81 Kb
↓
Sample file: Empty Avro: schema only, zero records
Empty Avro: schema only, zero records
179 bytes
↓

🧠 How Avro works

Avro serializes data row by row, and the schema (in JSON) is stored right in the file header — the reader always knows the structure and needs no external description. Records are split into blocks, and each block is compressed with a codec (null, deflate, snappy). The format is built for streaming writes and schema evolution.

📄 JSON schema types and fields 🗃️ header schema in file 🧱 Row-based records data blocks 🗜️ Codec null / deflate / snappy 💾 .avro ready file