Compare Parquet Files Online
Compare two Parquet files directly in your browser. See rows only in file A, only in file B, and rows common to both — no upload required.
How to compare Parquet files
- Drop the older file into the File A box and the newer file into the File B box. Each box shows the row and column count once the file has loaded.
- Click Compare. The tool first lists every column with a coloured dot: green for columns that match by name and type, yellow for a type mismatch, and blue or purple for columns found in only one file.
- Read the three summary cards: Only in File A, Only in File B, and In both files. Each card shows the full row count. Click a card or tab to preview up to 500 rows of that set.
- Click Download while a tab is active to save that set as a new file in the same format, for example prices_only_in_file_a.csv.
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.
Worked example
A supplier sends a new price list each month. You want to see which lines were dropped or repriced since last month, and which products are new.
Input (Parquet)
Parquet file (binary, columnar) — shown as a table with its schema
| sku | product | price |
|---|---|---|
| KB-101 | Wireless keyboard | 49.99 |
| MS-220 | Ergonomic mouse | 34.5 |
| HD-310 | USB-C hub | 59.99 |
| CB-015 | HDMI cable 2m | 12.99 |
Schema: sku VARCHAR, product VARCHAR, price DOUBLE
Settings
- File A: last month's price list (shown above)
- File B: this month's list. KB-101 and HD-310 are unchanged, MS-220 is now 32.0, CB-015 is gone, and DS-400 Laptop stand at 44.95 is new
- Active tab: Only in A
Result
| sku | product | price |
|---|---|---|
| MS-220 | Ergonomic mouse | 34.5 |
| CB-015 | HDMI cable 2m | 12.99 |
Rows are compared on every shared column at once, so a repriced product counts as a different row. MS-220 appears in Only in A with its old price and in Only in B with its new price. Only in B also holds DS-400. In both holds KB-101 and HD-310. Row order within each set is not guaranteed.
Working with Parquet files
Parquet stores a physical type for every column, and the comparison only uses columns whose name and type are identical in both files. Two exports of the same table from different tools often disagree here. An INT32 column in one file reads as INTEGER, an INT64 column in the other reads as BIGINT, and that column is dropped from the match. The same happens when one writer uses DECIMAL(10,2) and another uses DOUBLE for a price.
Timestamps are compared at their full stored precision. A file written with microsecond timestamps and one written with nanosecond timestamps have different column types, so check the schema panel before trusting the counts. Nested struct and list columns take part in the comparison when their full type matches, and are compared element by element. The three result sets are written as Parquet files with the column types of the compared columns preserved.
Frequently Asked Questions
Why is a Parquet column ignored when comparing two versions of the same table?
Its type differs between the files, for example INTEGER in one and BIGINT in the other, or a different timestamp precision. Only columns with the same name and the same type are used. The schema panel shows these in yellow.
Can I compare Parquet files that have nested columns?
Yes, as long as the struct or list has the same shape in both files. A struct that gained an extra field in one file counts as a different type and is left out of the comparison.
How does the comparison decide two rows are the same?
It uses only the columns that exist in both files with the same name and type, and treats two rows as equal when every one of those values matches. Rows only in A are found with a set difference, and rows in both with a set intersection. A changed value makes the row appear once in each Only set.
What happens to duplicate rows?
Each result set lists distinct rows. If a row appears three times in File A and never in File B, it appears once in Only in A. Use Find Duplicates first if the number of copies matters.
Can I match rows on a key column and see which fields changed?
No. The tool compares whole rows, not rows joined on an ID. To see field-level changes, load both files in the SQL workspace and join them on the key column.
Related Tools
Compare Schema of Parquet Files Online
Compare the schemas of two Parquet files directly in your browser. See which columns exist in each file and spot type mismatches — no upload required.
Deduplicate Parquet Files Online
Remove duplicate rows from Parquet files instantly in your browser. No upload, no server — 100% private.
Merge Parquet Files Online
Merge and concatenate multiple Parquet files into one, directly in your browser.
Compare CSV Files Online
Compare two CSV files directly in your browser. See rows only in file A, only in file B, and rows common to both — no upload required.
Compare Excel Files Online
Compare two Excel files directly in your browser. See rows only in file A, only in file B, and rows common to both — no upload required.
Compare JSON Files Online
Compare two JSON files directly in your browser. See rows only in file A, only in file B, and rows common to both — no upload required.
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.
Convert Parquet to CSV Online
Convert Parquet files to CSV format directly in your browser. No upload required — your data never leaves your device.