Convert Parquet to Arrow Online
Convert Parquet files to Arrow format directly in your browser. No upload required — your data never leaves your device.
About converting Parquet to Arrow
Parquet and Arrow are both columnar, but they solve different problems. Parquet is a storage format: encoded, compressed and split into row groups so that it is small on disk and cheap to scan from S3. Arrow is the in-memory layout that pandas 2, Polars, DuckDB and Spark use while they compute. An Arrow IPC file is that memory layout written to disk, so a reader can memory-map it and start work without decoding anything.
People convert Parquet to Arrow to feed tools that read IPC natively: a Polars or pyarrow script that memory-maps a file many times a day, a JavaScript app using the apache-arrow package, Perspective or other browser dashboards, or a service that sends record batches over Arrow Flight. For a file you reload often, skipping Parquet decompression on every read can be worth the extra disk space.
The output is the Arrow IPC file format, the same container as Feather version 2. Buffers are written uncompressed, so expect the file to be several times larger than the Snappy-compressed Parquet it came from. Parquet-specific metadata, such as the pandas schema block or row group statistics, is not copied.
Not sure which format you need? Read the Apache Arrow vs Parquet comparison.
Worked example
A small orders export with an ID, a customer name, a date, an amount (one missing) and a true/false flag, converted with the default settings.
Input (Parquet)
Parquet file (binary, columnar) — shown as a table with its schema
| order_id | customer | order_date | amount | shipped |
|---|---|---|---|---|
| 1001 | Acme Ltd | 2026-03-02 | 249.5 | true |
| 1002 | Brightside Co | 2026-03-02 | 1200 | false |
| 1003 | Acme Ltd | 2026-03-05 | 89.99 | true |
| 1004 | Northwind | 2026-03-07 | NULL | false |
Schema: order_id BIGINT, customer VARCHAR, order_date DATE, amount DOUBLE, shipped BOOLEAN
Output (Arrow)
Arrow IPC file (binary, columnar) — shown as a table with its schema
| order_id | customer | order_date | amount | shipped |
|---|---|---|---|---|
| 1001 | Acme Ltd | 2026-03-02 | 249.5 | true |
| 1002 | Brightside Co | 2026-03-02 | 1200 | false |
| 1003 | Acme Ltd | 2026-03-05 | 89.99 | true |
| 1004 | Northwind | 2026-03-07 | NULL | false |
Schema: order_id BIGINT, customer VARCHAR, order_date DATE, amount DOUBLE, shipped BOOLEAN
What changes when you convert Parquet to Arrow
- The file starts with the ARROW1 magic bytes. pyarrow.feather.read_table, pandas.read_feather, polars.read_ipc and pyarrow.ipc.open_file all read it.
- Types map directly: BIGINT becomes int64, DOUBLE becomes float64, DECIMAL(p,s) becomes decimal128(p,s), DATE becomes date32, TIMESTAMP becomes timestamp in microseconds, and VARCHAR becomes utf8.
- Timezone-aware timestamps stay timezone-aware, and NULLs are kept in the validity bitmap rather than turned into empty strings.
- Structs, lists and maps stay as Arrow struct, list and map types. Unlike the text outputs, DECIMAL, DATE and TIMESTAMP values keep their types instead of being written as strings.
- Record batch sizes are chosen by the converter. They do not follow the row groups of the source Parquet file.
Your file is processed locally in your browser and is never uploaded. The free limit is 50 MB per file; larger files work if your device has the memory for them.
Frequently Asked Questions
Is the output a Feather file?
Yes, in the sense that Feather version 2 is the Arrow IPC file format. Rename it to .feather if a tool expects that extension. It is not Feather version 1, the older format that pandas wrote before 2020.
Why is the Arrow file so much bigger than my Parquet file?
Parquet applies dictionary and run-length encoding plus Snappy or Zstandard compression. The Arrow output stores raw, uncompressed buffers so that readers can use them in place. Three to ten times larger is normal for data with many repeated values.
Should I store data long term as Arrow instead of Parquet?
Usually not. Parquet is smaller, is read by every warehouse and lake engine, and has row group statistics for skipping data. Keep Parquet as the stored copy and produce Arrow when a process needs fast repeated loads.
What is Arrow format?
Apache Arrow is a columnar in-memory format designed for zero-copy reads and high-speed exchange between data systems.
Related Tools
Convert Arrow to Parquet Online
Convert Arrow files to Parquet format directly in your browser. No upload required — your data never leaves your device.
Arrow Viewer Online
View and inspect Arrow files directly in your browser. Browse rows, check column names and data types — no upload required, your data stays on your device.
Parquet Viewer Online
View and inspect Parquet files directly in your browser. Browse rows, check column names and data types — no upload required, your data stays on your device.
Validate Parquet Files Online
Validate Parquet file structure in your browser. Check null counts, distinct values, and data types for every column — no upload required.
Deduplicate Parquet Files Online
Remove duplicate rows from Parquet files instantly in your browser. No upload, no server — 100% private.
Convert Parquet to CSV Online
Convert Parquet files to CSV format directly in your browser. No upload required — your data never leaves your device.
Convert Parquet to Excel Online
Convert Parquet files to Excel format directly in your browser. No upload required — your data never leaves your device.
Convert Parquet to JSON Online
Convert Parquet files to JSON format directly in your browser. No upload required — your data never leaves your device.