Read, write, and generate spreadsheet content with AI. Create data from natural language prompts and automate your spreadsheet workflows.
The Google Sheets integration enables you to work with spreadsheets directly from ARKA AI. List your sheets, read data from specific ranges, write data programmatically, and even generate entire datasets from natural language descriptions.
Browse all Google Sheets in your Drive:
Read data from any sheet or range in your spreadsheets:
Browse or search for the spreadsheet you want to read from.
Enter the range you want to read using A1 notation (e.g., A1:D10, Sheet1!A:B, or just the sheet name for all data).
Select how values should be rendered: Formatted (as displayed) or Raw (underlying values).
Data is displayed in a table view. You can copy, export, or use it in your AI conversations.
| Notation | Description |
|---|---|
| A1:D10 | Cells from A1 to D10 in the first sheet |
| Sheet1!A1:D10 | Cells from A1 to D10 in 'Sheet1' |
| A:D | All rows in columns A through D |
| 1:10 | All columns in rows 1 through 10 |
| Sheet2!A:A | All of column A in 'Sheet2' |
| 'My Sheet'!A1:B5 | Range in a sheet with spaces in the name |
Read multiple ranges at once for efficiency. Provide a comma-separated list of ranges:
A1:B10, D1:E10, Sheet2!A1:A20Write data to any cell or range in your spreadsheets:
Choose the spreadsheet you want to write to.
Enter the cell or range where data should be written (e.g., A1 or Sheet1!C5).
Either enter data manually or use AI generation (see below) to create content from a prompt.
Review the data preview and click Write to update the spreadsheet.
Overwriting Data
Writing to a range will overwrite any existing data in those cells. Make sure you're targeting the correct range before confirming.
Generate spreadsheet data from natural language prompts. Describe the data you need, and the AI will create structured content ready to insert.
In the Sheets view, select a spreadsheet and click "Generate with AI."
Write a natural language description of the data you want. Be specific about columns, data types, and number of rows.
Enter where the generated data should be written (e.g., A1).
Preview the generated data, make any adjustments, then confirm to write it to the spreadsheet.
Here are effective prompts for generating spreadsheet content:
Create a table of 20 fictional product sales with columns:
- Product Name
- Category (Electronics, Clothing, Home, Food)
- Quantity Sold (10-500)
- Unit Price ($5-$200)
- Total Revenue
- Sale Date (last 30 days)Generate a contact list with 15 rows containing:
- Full Name
- Email Address
- Phone Number
- Company
- Job Title
- CityCreate a project tracking table with 10 tasks:
- Task Name
- Assignee
- Priority (High/Medium/Low)
- Status (Not Started/In Progress/Complete)
- Due Date (next 2 weeks)
- Estimated HoursGenerated data is returned as a JSON array of arrays, which maps directly to spreadsheet rows and columns:
[
["Name", "Email", "Department"],
["John Doe", "john@example.com", "Sales"],
["Jane Smith", "jane@example.com", "Marketing"]
]All write operations are logged for reference:
Be Specific in Prompts
Include column names, data types, value ranges, and row count for more accurate generation.
Test with Small Ranges
When trying new prompts, generate a small dataset first to verify the format before creating larger datasets.
Include Headers
Ask for column headers in your prompt so the generated data is self-documenting.
Back Up Important Data
Before writing to sheets with important data, consider duplicating the sheet as a backup.