Data Format Converter
Convert between JSON, YAML, XML, CSV, and Parquet — right in your browser.
Overview
How to convert between JSON, YAML, XML, CSV and Parquet
Convert structured data between JSON, YAML, XML, CSV, and Apache Parquet without uploading anything. Drop in a file and pick the output format — conversion happens locally in your browser. The source format is detected from the file extension.
Use cases
When to convert between JSON, YAML, XML, CSV and Parquet
- 01
Converting a YAML config to JSON for an API
- 02
Turning a CSV export into JSON records
- 03
Reformatting XML into readable YAML
Examples
Data converter examples
config.yaml
ex 01config.json
Convert a YAML config to JSON
data.csv
ex 02data.json
Convert a CSV table to JSON records
data.csv
ex 03data.parquet
Convert a CSV table to Apache Parquet
FAQ
Data converter: common questions
No. Conversion runs entirely in your browser using local JavaScript — your files never leave your device.
From the file extension (.json, .yaml/.yml, .xml, .csv, .parquet). Rename the file if its extension doesn't match its contents.
CSV is a flat table with no way to express nesting, so nested objects and arrays have to be flattened into columns or serialised into a single cell. If your JSON is deeply nested, expect to lose its shape — YAML, XML or Parquet preserve structure and CSV cannot. For a flat array of objects, the conversion is clean: each key becomes a column.
No. JSON has no comment syntax, so anything you annotated in the YAML is dropped permanently. If the comments matter, keep the YAML as your source of truth and treat the generated JSON as a build artefact you can regenerate.
Parquet stores data by column rather than by row, which makes reading a few columns out of a wide table dramatically faster and compresses far better because similar values sit together. Expect roughly 5-10× smaller files than the equivalent CSV, plus a real typed schema. The trade-off is that it is binary — you cannot open it in a text editor.
JSON numbers are IEEE 754 doubles, so integers beyond 2^53 lose precision silently — this is a property of JSON itself, not of this converter. If you are moving IDs or timestamps that exceed that range, carry them as strings, or use Parquet, which has real integer types.
More
Tools that pair well
Last updated