Avro to JSON Converter
Paste an Apache Avro Object Container File (.avro) as base64 or hex and read it as JSON. The schema is embedded in the file, so nothing else is needed. Choose a JSON array, NDJSON, or a full view with the writer schema. Runs entirely in your browser — no server, no sign-up.
Read Apache Avro files as JSON
Apache Avro is a compact, schema-driven binary format used widely in data
pipelines (Kafka, Hadoop, Spark). An Object Container File (.avro, often
called an OCF) bundles the records together with the writer schema that
describes them — so the file is fully self-describing. This tool reads that
container and turns the records back into plain JSON, right in your browser.
Paste the file's bytes as base64 or hex and pick how you want the
output. No .avsc schema file is required, because the schema travels inside
the container.
Output formats
- Records — a pretty-printed JSON array of every record in the file. Best for reading or piping into another JSON tool.
- NDJSON — newline-delimited JSON, one compact record per line. Handy for streaming into log tools or loading row-by-row.
- Full — an object with the embedded writer schema, the record count, and the records. Use it when you want to see exactly which schema the file was written with.
How values are decoded
- Avro logical types are unwrapped: dates and timestamps come through as their
underlying integers, a
uuidbecomes its string form, anddecimal,bytes, andfixedvalues are base64-encoded so the raw bytes survive. - Unions are flattened to the actual branch value, and records, arrays, and maps map onto JSON objects and arrays directly.
Common uses
- Inspect an
.avrofile dumped from a Kafka topic or a data lake. - Convert Avro records to JSON or NDJSON for quick
jq-style exploration. - Confirm which schema an Avro file was actually written with.
This tool reads container files that embed their schema — not bare, single-object-encoded Avro values, which carry no schema on their own. Everything runs locally in your browser via WebAssembly — your data never leaves your machine.
FAQ
Do I need to supply the .avsc schema file?
No. An Avro Object Container File embeds its writer schema in the file header,
so the tool reads it from the bytes you paste. That's also why bare,
single-object-encoded Avro values are rejected — they carry no schema. If you
see the error "not an Avro Object Container File", the bytes are missing the
Obj\x01 magic that every OCF starts with.
How should I encode the file bytes — base64 or hex?
Either works. With encoding set to auto (the default), input made only of
hex digits with an even length is treated as hex; anything else is decoded as
base64. Base64 may use the standard or URL-safe alphabet and padding is
optional; hex may contain spaces, : or - separators, which are ignored. On
a shell, base64 < file.avro produces ready-to-paste input.
Why do dates, timestamps, and decimals come out as numbers or base64 strings?
Logical types are unwrapped to their underlying representation: date,
time, and timestamp values appear as their raw integers, a uuid becomes
its usual string form, and decimal, bytes, and fixed values are
base64-encoded so no raw bytes are lost. Unions are flattened to the actual
branch value.
How can I see which schema the file was written with?
Pick the Full output format. It returns an object with the embedded
writer schema, the record count, and the decoded records — useful for
confirming what a producer actually wrote to a Kafka topic or data lake.
Developer & Automation Access
Run it from the terminal
Same engine as this page, headless — via the gizza CLI:
gizza tool avro-to-json "T2JqAQQ..."New to the CLI? Get gizza →
Open it by URL
Pre-fill and auto-run this tool with query parameters — the names match the API/CLI:
https://gizza.ai/tools/avro-to-json/?input=T2JqAQQ...&encoding=auto&format=recordsMachine-readable descriptor: tool.json — title + parameters JSON Schema for agents.