Data

XLSX vs CSV: When to Use Each Format

Published Apr 13, 2026 6 min read By ChangeThisFile Team

Quick Answer

XLSX is a rich spreadsheet format that stores formulas, cell formatting, charts, multiple sheets, named ranges, conditional formatting, and data validation. CSV is plain text — it stores only values (no formulas, no formatting, no multiple sheets). Use XLSX when you need to preserve spreadsheet functionality or share data within the Office/Sheets ecosystem. Use CSV for data exchange: importing into databases, feeding ML pipelines, and any situation where the consuming system needs to parse the data programmatically.

Quick Verdict

Best for Data Pipelines

CSV

Every database, ML library, and analytics tool natively imports CSV — universal programmatic compatibility

Best for Spreadsheet Users

XLSX

Preserves formulas, formatting, charts, and multiple sheets that CSV cannot represent

Smallest for Large Data

XLSX

ZIP compression and shared string deduplication make XLSX 50-75% smaller than CSV for datasets over 1,000 rows

Fastest to Parse

CSV

pandas reads CSV 3-10x faster than XLSX — plain text streaming requires no decompression or XML parsing

Benchmarks

Test data: 10,000-row dataset with 5 columns: ID (integer), name (text), address (text with commas), date (ISO), amount (decimal)

format	file Size	parse Time	formula Support	formatting Support
CSV	820 KB	120 ms (pandas)	None	None
XLSX	340 KB	850 ms (pandas + openpyxl)	Full (computed and stored)	Full (fonts, colors, number formats)

Evaluation Criteria

Tool Compatibility ●●●●●

How many databases, programming languages, and data tools can import the format natively

Data Richness ●●●●○

Support for formulas, formatting, multiple sheets, charts, and data validation

File Size ●●●○○

Storage footprint for equivalent data content

Parse Performance ●●●●○

Speed of reading and processing in data analysis workflows

End-User Accessibility ●●●●●

How easily non-technical users can open, view, and work with the file

XLSX and CSV both store tabular data, but they represent fundamentally different philosophies about what a data file should contain. CSV is brutally simple: rows of values, columns separated by commas, nothing else. XLSX is a full spreadsheet application format — a ZIP archive containing XML files that encode not just data values but the entire computational and visual state of a spreadsheet.

This difference is not a matter of features — it's a matter of purpose. CSV is designed to move data between systems. XLSX is designed to be a complete working environment for a human spreadsheet user. Choosing the wrong format for your situation means either losing critical information (CSV when you needed formulas) or creating unnecessary complexity (XLSX when a database just needs the values).

What Each Format Stores

Understanding what XLSX contains — versus what it loses when you export to CSV — is the starting point for the right format choice.

What XLSX Stores

XLSX (Office Open XML Spreadsheet) is a ZIP archive. Unzip an .xlsx file and you'll find XML files inside. It stores:

Cell values. The actual data: text, numbers, dates, booleans.
Formulas. =SUM(B2:B100), =VLOOKUP(A2, Sheet2!A:B, 2, FALSE), pivot table calculations. These are stored as formula strings, re-evaluated when you open the file.
Cell formatting. Font, size, bold/italic, color, borders, fill color, number formats ($#,##0.00, yyyy-mm-dd, percentage).
Multiple sheets. A single .xlsx file can contain dozens of worksheets, each with its own data and formatting.
Charts and graphs. Bar charts, line charts, pivot charts — stored as embedded chart objects.
Named ranges. TaxRate that refers to $B$2 — formula references that are human-readable.
Data validation. Dropdown lists, numeric constraints, date ranges — the rules that prevent invalid data entry.
Conditional formatting. Highlight rules — cells that turn red when below a threshold, color scales, data bars.
Pivot tables. Summary aggregations with filters, groupings, and calculated fields.
Merged cells. Cells spanning multiple rows or columns — common in reports and headers.
Comments and notes. Reviewer annotations attached to specific cells.
Shared strings. XLSX deduplicates repeated string values in a shared string table for file size efficiency.

What CSV Stores (and What It Loses)

CSV stores exactly one thing: the computed values of cells. When you export an XLSX to CSV:

Formulas → computed values. =SUM(B2:B100) becomes the number 4827.50. The formula is gone. If the source data changes, the CSV won't update.
Formatting → lost. Bold headers, currency formatting, color coding — none of it survives. Every cell is a plain string in CSV.
Multiple sheets → one sheet. CSV can only represent a single table. When you export a multi-sheet XLSX to CSV, you either export one sheet or create multiple CSV files.
Charts → lost. No chart format exists in CSV. Charts are visual representations of the data; CSV keeps the data only.
Number formats → string representation. A date stored as a serial number with a date format in XLSX becomes either a date string (2026-04-13) or an Excel serial number (46384) in CSV, depending on how the export is performed. This is a frequent source of data corruption.
Data types → ambiguous. CSV has no type system. The number 007 may be interpreted as integer 7 when opened in Excel. The date 3/5 might be parsed as March 5, May 3, or the fraction 0.6 depending on locale and software.

The rule of thumb: CSV preserves the what but not the how. It's the data without the context, computation, or presentation.

File Size Comparison

For purely numeric or simple tabular data, CSV is smaller than XLSX. For complex spreadsheets with formatting, XLSX's ZIP compression makes the comparison less clear.

Data Type	CSV Size	XLSX Size	Notes
10,000 rows of numbers (5 columns)	~450 KB	~180 KB	XLSX ZIP compression wins on repetitive numeric data
10,000 rows with text (mixed)	~800 KB	~350 KB	XLSX shared strings table deduplicates repeated values
Simple 100-row table, no formatting	~5 KB	~8 KB	XLSX overhead from XML structure and required files
Complex spreadsheet with charts, formatting	N/A (can't represent)	~2-20 MB	Charts and embedded objects dominate size

Counterintuitively, XLSX is often smaller than CSV for large datasets because XLSX uses ZIP compression internally. A 1 million-row CSV file might be 150MB; the equivalent XLSX could be 30-50MB. However, XLSX has baseline overhead from its XML structure — for small files (under ~200 rows), a plain CSV is usually smaller due to the XML scaffolding required even for empty XLSX files.

For sharing large datasets via email or messaging apps with file size limits, this matters. Convert to whichever is smaller: XLSX to CSV for simple tabular data, or keep XLSX for complex spreadsheets where the compression benefit outweighs the format overhead.

Compatibility Across Tools and Systems

Compatibility is the primary reason to choose CSV over XLSX in many data workflows:

CSV: Near-Universal Compatibility

Every programming language has a CSV parser. Every database can import CSV. Every BI tool, data warehouse, and analytics platform accepts CSV. It's plain text — any text editor can open it, any scripting language can process it without a library, and any system that can read files can parse it.

Database imports: PostgreSQL COPY FROM, MySQL LOAD DATA INFILE, BigQuery load jobs, Snowflake COPY INTO — all accept CSV.
Programming languages: Python's built-in csv module, pandas read_csv, R's read.csv, Go's encoding/csv, Node's csv-parse — zero external dependencies.
Data tools: Jupyter notebooks, dbt, Spark, DuckDB, Hadoop — all treat CSV as a first-class input format.
Legacy systems: Systems from the 1980s and 1990s that predate Excel's XLSX format can still read CSV. XLSX requires modern software.

XLSX: Spreadsheet-Ecosystem Compatibility

XLSX compatibility is excellent within the spreadsheet ecosystem but less universal for programmatic processing:

Excel, Google Sheets, LibreOffice Calc: Native format support, full fidelity (with some compatibility notes between vendors for advanced features).
Programming languages: Requires a library — openpyxl or xlrd in Python, SheetJS (xlsx) in JavaScript, Apache POI in Java. These libraries are mature and widely used but add a dependency.
Cloud services: Google Drive, Dropbox, OneDrive all preview XLSX natively. Many SaaS platforms can import XLSX (HubSpot, Salesforce, Shopify).
Databases: Databases don't natively import XLSX. You need an intermediate step: open in a spreadsheet app, export to CSV, then import. Some ETL tools handle XLSX directly but it's never a native database format.

The practical rule: if the data is going into a programming pipeline, database, or analytics tool, convert to CSV first. If the data is going to a human who will use it in a spreadsheet application, keep XLSX.

When to Convert Between XLSX and CSV

The most common conversion scenarios:

XLSX → CSV for database import. You have a spreadsheet of customer records, product inventory, or financial data that you need to import into a database or data warehouse. Export each sheet as CSV and use the database's bulk import tool. This is the most common XLSX-to-CSV use case. Use ChangeThisFile's XLSX to CSV converter.
XLSX → CSV for Python/pandas analysis. Pandas can read XLSX directly (pd.read_excel), but pd.read_csv is 3-10x faster for large files because CSV parsing is simpler. For one-time analysis, read XLSX directly. For repeated processing of large files, convert to CSV first.
CSV → XLSX for sharing reports. You have a CSV data export that you want to format as a professional report — add headers, apply number formatting, create summary rows. Import the CSV into Excel or Google Sheets, apply formatting, and save as XLSX. Use CSV to XLSX conversion to start.
XLSX → TSV for data pipelines that need tab-delimited output. If your data contains commas (common in XLSX — addresses, descriptions, formatted numbers), exporting to TSV instead of CSV avoids quoting issues. Use XLSX to TSV conversion.
XLSX → CSV for ML training data. Machine learning pipelines typically expect flat CSV or TSV. Export the relevant sheet, ensure numeric values don't have thousands separators or currency symbols, and check that dates are in ISO format (YYYY-MM-DD) rather than locale-specific formats.

How to Convert Without Excel

You don't need Excel or Google Sheets to convert between XLSX and CSV. Several options work without any spreadsheet application:

Browser-Based Conversion (No Software Required)

ChangeThisFile's converters run entirely in your browser using JavaScript — your file never leaves your device. Convert XLSX to CSV: XLSX to CSV. For the reverse: CSV to XLSX. This is the fastest option for one-off conversions — no software installation, no file upload.

Python (pandas + openpyxl)

For programmatic or batch conversion:

# XLSX to CSV (all sheets)
import pandas as pd

xl = pd.ExcelFile('input.xlsx')
for sheet_name in xl.sheet_names:
    df = xl.parse(sheet_name)
    df.to_csv(f'{sheet_name}.csv', index=False)

# CSV to XLSX
df = pd.read_csv('input.csv')
df.to_excel('output.xlsx', index=False, engine='openpyxl')

Requires: pip install pandas openpyxl. For read-only XLSX parsing (faster, no write support): pip install xlrd. Note that xlrd 2.0+ dropped support for XLSX (it now only reads XLS) — use openpyxl for XLSX.

LibreOffice Command Line

LibreOffice Calc can convert headlessly:

# XLSX to CSV (converts the first sheet)
libreoffice --headless --convert-to csv input.xlsx

# Multiple files
libreoffice --headless --convert-to csv *.xlsx

Limitations: LibreOffice's headless CSV export only handles the first sheet and uses the locale's default CSV settings (delimiter, encoding). For multi-sheet XLSX, you'll need pandas or a dedicated tool.

The XLSX vs CSV choice is usually obvious once you know who or what is consuming the data. Human spreadsheet users need XLSX — they need formulas, formatting, and the ability to see and edit all sheets. Databases, data pipelines, programming scripts, and ML tools need CSV — they need plain values they can parse without a spreadsheet engine.

When you need to switch: convert XLSX to CSV for data pipelines and database imports, or convert CSV to XLSX to bring data into a spreadsheet for formatting and analysis. For data containing commas, XLSX to TSV avoids the quoting complexity of CSV. All conversions run in the browser — no Excel, no uploads.

Key Takeaways

XLSX stores formulas, formatting, multiple sheets, charts, and pivot tables — CSV stores only the computed values with no formatting
CSV is universally compatible — every database, programming language, and analytics tool accepts it; XLSX requires a library or application
Counterintuitively, XLSX is often smaller than CSV for large datasets because XLSX uses ZIP compression and a shared string table
Excel opens .csv files with auto-detection; when Excel opens a CSV, it may corrupt leading zeros, reformat dates, and apply locale-specific number parsing
XLSX formulas are lost permanently when you export to CSV — the export captures the computed value at export time, not the formula
For database imports and data pipelines, always convert to CSV first — no major database natively imports XLSX
Converting without Excel is easy — ChangeThisFile's browser converter, pandas in Python, or LibreOffice CLI all handle XLSX/CSV conversion without Microsoft Office

Frequently Asked Questions

What is the difference between XLSX and CSV?

XLSX is a full spreadsheet format that stores formulas, cell formatting, multiple worksheets, charts, pivot tables, and data validation. CSV is plain text — it stores only cell values with no formatting, formulas, or multiple sheets. When you export XLSX to CSV, formulas are replaced by their computed values and all formatting is lost. CSV is universally compatible (works in any database, programming language, or tool). XLSX requires a spreadsheet application or library to open.

How do I convert XLSX to CSV without Excel?

Three options: (1) ChangeThisFile's XLSX to CSV converter runs in your browser — drop the file, download the CSV, no software needed. (2) Python with pandas: import pandas as pd; pd.read_excel('file.xlsx').to_csv('file.csv', index=False) — requires pip install pandas openpyxl. (3) LibreOffice CLI: libreoffice --headless --convert-to csv file.xlsx — LibreOffice must be installed but it's free. All three work without Microsoft Office.

Does converting XLSX to CSV lose data?

It loses formatting and formulas, but not values. When you export XLSX to CSV: formulas are replaced by their computed values at export time, cell formatting (bold, colors, number formats) is stripped, charts and images are dropped, multiple sheets become multiple CSV files or you choose one sheet, and data validation rules are removed. The raw data values in cells are preserved. If you need the formulas, keep the XLSX — there's no way to recover a formula from a CSV.

Which is better for large datasets — XLSX or CSV?

For large datasets going into a database or data pipeline, CSV is better because it's universally supported and pandas reads CSV 3-10x faster than XLSX. For storage and sharing, XLSX can actually be smaller than CSV for large datasets because XLSX uses ZIP compression internally and deduplicates repeated string values — a 1 million-row CSV might be 150MB while the equivalent XLSX is 30-50MB. For raw programmatic processing speed, CSV wins. For file size, XLSX often wins on large data.

How do I import an XLSX file into a database?

Most databases don't natively import XLSX — you need to convert to CSV first. Process: (1) Export each relevant sheet as CSV from Excel/Google Sheets, or use Python/pandas to convert programmatically. (2) Clean the CSV — remove header rows, ensure consistent formatting, check for leading zeros preserved as strings. (3) Use the database's bulk import: PostgreSQL COPY FROM, MySQL LOAD DATA INFILE, BigQuery bq load, Snowflake COPY INTO. Always check encoding (use UTF-8) and delimiter settings.

Why does Excel mess up my CSV when I open it?

Excel applies type inference when opening CSV files — it guesses the type of each column and reformats values accordingly. Common damage: (1) Leading zeros stripped — ZIP codes, phone numbers, IDs like 007 become 7. (2) Dates reformatted — 2026-04-13 may become 4/13/2026 or an Excel serial number. (3) Large numbers converted to scientific notation — 1234567890123 becomes 1.23457E+12. (4) Special values misinterpreted — 3/5 reads as a date, TRUE becomes boolean. Fix: use Data → From Text/CSV with explicit column type settings rather than double-clicking the file.

Can XLSX store multiple tables like a database?

XLSX can store multiple sheets (tabs), each containing one table. This is structurally different from a relational database — there's no foreign key enforcement, no query language, and no referential integrity. It's possible to have a Customers sheet and an Orders sheet with a shared CustomerID column and use VLOOKUP to join them, but this is error-prone at scale. For actual multi-table data with relationships, a database (SQLite, PostgreSQL) or a format like JSON is more appropriate.

What happens to formulas when I save as CSV?

Formulas are permanently replaced by their computed values at the time of export. If cell B2 contains =A2*1.2 and A2 is 100, the CSV will contain 120 — there's no way to tell from the CSV that a formula was involved. The formula is gone. If you need to preserve formulas, keep the XLSX file. There is no way to recover a formula from a CSV. This is the primary reason to keep XLSX as the source of truth and export CSV only for downstream consumption.

How do I convert CSV to XLSX and keep the formatting?

CSV has no formatting to preserve — it's plain values. When you convert CSV to XLSX, you get an XLSX with unformatted cells (default font, no colors, no number formats). You have to apply formatting after the conversion. To convert with Python and add basic formatting: use openpyxl to write the XLSX, then apply styles programmatically. For a quick no-code conversion with subsequent manual formatting: use ChangeThisFile's CSV to XLSX converter to create the XLSX, then open it in Excel or Google Sheets and apply formatting.

Compare Formats

CSV vs XLSX

Ready to convert your files?

Use ChangeThisFile to convert between 600+ formats — free, fast, and private.

Start Converting