Convert Parquet to Excel (XLSX)
A native browser-based .xlsx converter for Apache Parquet — preserving column types, dates, and large datasets up to Excel's 1,048,576-row limit. The online tool is in development; meanwhile use the Parquet → CSV converter (Excel opens CSV natively) or one of the methods below.
Why XLSX instead of CSV?
CSV is the path of least resistance — Excel opens it directly. But XLSX has real advantages when the receiving user is going to do analysis, not just a one-off look:
- Real types. Numbers stay numeric, dates stay dates. Excel won't auto-mangle
"001234"into1234. - Multiple sheets. One workbook can hold metadata, schema, and data on separate tabs.
- Formatting and formulas. Number formats, conditional formatting, freeze panes — none of which CSV preserves.
- UTF-8 by default. No BOM dance, no "Excel mangled my Cyrillic / Chinese / emoji".
The downside: Excel has a hard limit of 1,048,576 rows per sheet. Datasets larger than that need to be split, sampled, or summarised before export.
Convert Parquet to XLSX manually (until the online tool ships)
Python (pandas + openpyxl)
pip install pandas openpyxl pyarrow
import pandas as pd
df = pd.read_parquet("data.parquet")
df.to_excel("data.xlsx", sheet_name="data", index=False)DuckDB (one-liner)
duckdb -c "INSTALL excel; LOAD excel;
COPY (SELECT * FROM 'data.parquet')
TO 'data.xlsx' (FORMAT 'xlsx', HEADER)"Both methods load the entire Parquet file into memory before writing the .xlsx — fine for files under a few hundred MB.
In the meantime — use the CSV converter
The fastest browser-based path right now is to convert to CSV and open the CSV in Excel. Excel auto-detects the UTF-8 BOM that the Parqui CSV converter prepends, so unicode characters render correctly.
Open Parquet → CSV converter →