How to Set Up Quadratic for Spreadsheet Analysis

What Quadratic does and why it matters

Quadratic is a browser-based spreadsheet that combines traditional formulas with Python, SQL, and JavaScript in the same workbook. If you have ever hit the limits of Excel or Google Sheets when trying to run a pandas operation, connect to a database, or automate a repetitive cleaning step, Quadratic fills that gap without forcing you into a full IDE.

The platform is free for individual use, runs entirely in Chrome, and supports real-time collaboration. It is used by teams at Bloomberg, Google, and MIT, and currently has over 250,000 users. This guide walks through how to set up Quadratic from scratch and run a basic analysis on your own data.

Step 1: Create your account

Go to app.quadratichq.com and sign up with a Google account or email. There is no software to install. Once you log in, you land on a dashboard that shows your recent files and a set of example templates.

Click "Create blank spreadsheet" to start fresh, or pick one of the example files to explore how Python and SQL cells work in practice. The example files are worth browsing first if this is your first time using a code-enabled spreadsheet.

Step 2: Import your data

Quadratic accepts CSV, Excel (.xlsx), and PDF files. To import, click the file menu in the top left and select "Import." Drag your file into the dialog or browse your local drive. The data loads directly into the sheet grid, with columns and rows mapped automatically.

If your data lives in a database, skip the file import. Quadratic connects natively to PostgreSQL, MySQL, and Snowflake. Open a new SQL cell (press / in any cell, then select "SQL"), configure your connection credentials in the panel that appears, and write your query. Results populate the sheet starting at the cell where you placed the SQL block.

For API-based data sources, you can use a Python or JavaScript cell to make HTTP requests and write the results back to the grid. Quadratic includes common libraries like pandas and requests in its Python runtime, so there is nothing extra to install.

Step 3: Use the AI assistant

The AI assistant is the fastest way to start analyzing data if you are not sure what code to write. Open it by clicking the sparkle icon in the toolbar or pressing the AI shortcut in the command palette.

Type a plain-English question like "Show me the top 10 customers by total spend" or "Clean up the date column and convert it to YYYY-MM-DD format." The assistant generates a Python or formula cell, places it in your sheet, and runs it. You can review the generated code, edit it, and re-run as needed.

Quadratic supports Claude 3.7 Thinking, Claude 3.5 Sonnet, and GPT-4o as backend models. The AI is not a black box. Every operation it performs is visible as editable code in your sheet, which means you can verify, modify, or learn from what it produces.

Step 4: Write code cells manually

If you prefer to write your own code, press / in any empty cell to open the command palette. Select Python, JavaScript, or SQL to open the inline code editor.

For Python, the runtime includes pandas, NumPy, and other standard data science libraries. You reference spreadsheet data by cell address. A single cell reference like cell(0, 0) returns a value. A range reference like cells((0, 0), (10, 3)) returns a DataFrame. Your code output is written back to the sheet starting at the cell position.

Here is a simple example that calculates a running total:

import pandas as pd

# Read column A (revenue) starting from row 1
df = cells((0, 1), (0, 100))
df.columns = ['revenue']
df['running_total'] = df['revenue'].cumsum()

# Output goes back to the sheet
df['running_total']

For SQL cells, the editor includes a schema browser on the right side that shows your connected tables and columns. Write your query, hit run, and results fill the grid.

Step 5: Create charts and visualizations

Quadratic includes a built-in charting tool. Select the data range you want to visualize, then click the chart icon in the toolbar. Choose from bar, line, scatter, pie, and other standard chart types.

You can also generate charts programmatically using Python. The Plotly library is available in the Python runtime, so you can create interactive visualizations directly in code cells. The chart renders inline in your spreadsheet.

For a quick approach, ask the AI assistant to "create a bar chart of monthly revenue from column B." It generates the appropriate code and renders the chart in your sheet.

Step 6: Collaborate and share

Quadratic supports real-time multiplayer editing, similar to Google Sheets. Click "Share" in the top right to invite teammates by email. Permission levels include viewer and editor roles.

Version history is built in. Click the clock icon to browse previous states of your spreadsheet and revert if needed. This is especially useful when multiple people are editing code cells simultaneously.

For presentation or reporting, you can export your sheet to CSV or take snapshots of specific views. If your data is connected via SQL, the sheet stays live and updates when you re-run your queries.

When Quadratic works better than Excel or Google Sheets

Quadratic is not a replacement for every spreadsheet workflow. For simple budgets, trackers, or forms, Excel and Google Sheets work fine. Where Quadratic shines is in scenarios where your analysis outgrows formulas.

If you find yourself writing nested INDEX-MATCH chains or VLOOKUP formulas that span multiple sheets, a Python cell can often do the same thing in three lines. If you need to pull data from a production database for a weekly report, a SQL cell eliminates the export-import loop. If you want to build a chart that updates automatically when new data arrives from an API, a JavaScript or Python cell handles that natively.

The AI assistant also changes the workflow for people who are not confident writing code. Instead of searching Stack Overflow for the right pandas syntax, you describe what you want and the assistant generates working code that you can inspect and modify. According to Quadratic, over 250,000 users have adopted this workflow, with teams at companies like Bloomberg and Apple among them.

Practical tips for getting the most out of Quadratic

Start with the AI assistant to explore unfamiliar datasets. It is particularly good at writing pandas code for cleaning and transformation tasks. Once you see the generated code, you can learn the patterns and start writing your own cells.

Use SQL cells for anything that involves filtering or joining data from a connected database. The query results land directly in your grid, which you can then process further with Python or formulas.

Keep your code cells organized. Each cell can be named and annotated, which makes complex workbooks easier to navigate when you return to them later.

If you are coming from Excel, the formula syntax is familiar. Standard functions like SUM, VLOOKUP, and IF work as expected. The difference is that you can drop into Python or SQL whenever a formula gets too complex, rather than building nested formula chains.

If you want to skip the setup process entirely for quick, one-off analyses, tools like VSLZ let you upload a file and get insights from a single prompt without configuring a spreadsheet environment.

Summary

Quadratic gives analysts and operators a way to combine spreadsheet simplicity with real code when they need it. The setup takes under two minutes: create an account, import your data, and start using AI, Python, SQL, or formulas to analyze it. The free tier is generous enough for individual use, and the collaboration features make it viable for small teams.