AI_TABLE
Generate a table of data from a natural language description
=AI_TABLE(prompt, [rows], [cols])Returns: string[][]
Overview
AI_TABLE generates structured tabular data from a plain English description, returning a formatted table that fills across rows and columns in your spreadsheet. Instead of manually researching and typing data into cells, simply describe the table you need — "Top 10 programming languages by popularity with year created and primary use case" — and AI_TABLE creates it instantly. The output includes headers and cleanly fills adjacent cells using Google Sheets' array spill behavior.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
prompt | string | Yes | A natural language description of the table you want to generate. Be specific about columns and content type. |
rows | number | No (5) | Number of data rows to generate (not counting the header row). |
cols | number | No (3) | Number of columns to generate. |
Examples
Generate a competitive analysis table
Creates a comparison table of email marketing platforms with specific columns.
=AI_TABLE("Compare the top 5 email marketing platforms with columns: Platform, Free Plan Limit, Starting Price, Best For", 5, 4)Output
| Platform | Free Plan Limit | Starting Price | Best For |
| Mailchimp | 500 contacts | $13/mo | Small businesses |
| ConvertKit | 1,000 subscribers | $15/mo | Creators |
| Brevo | 300 emails/day | $9/mo | Transactional email |
| ActiveCampaign | No free plan | $29/mo | Advanced automation |
| Klaviyo | 250 contacts | $20/mo | E-commerce |
Create a sample dataset for testing
Generates a realistic sample dataset for testing e-commerce formulas and charts.
=AI_TABLE("Realistic e-commerce order data with order ID, customer name, product, quantity, total price, order date", 10, 6)Output
| Order ID | Customer Name | Product | Quantity | Total Price | Order Date |
| ORD-001 | Maria Johnson | Wireless Mouse | 2 | $49.98 | 2025-01-15 |
| ORD-002 | David Chen | USB-C Hub | 1 | $34.99 | 2025-01-15 |
| ORD-003 | Sarah Williams | Mechanical Keyboard | 1 | $129.99 | 2025-01-16 |
Generate a content calendar structure
Creates an 8-week content calendar with topics and SEO targets.
=AI_TABLE("Content calendar for a fitness blog: week number, topic, target keyword, content type, target word count", 8, 5)Output
| Week | Topic | Target Keyword | Content Type | Target Word Count |
| 1 | 5 Beginner Gym Mistakes | gym mistakes beginners | Blog Post | 1,500 |
| 2 | Home Workout Equipment Guide | best home gym equipment | Buying Guide | 2,500 |
| 3 | Protein Intake Calculator | how much protein per day | Interactive Tool | 1,000 |
Build a project task breakdown
Generates a project plan with task estimates and dependencies for a website redesign.
=AI_TABLE("Website redesign project tasks with task name, responsible team, estimated hours, priority, dependencies", 12, 5)Output
| Task Name | Responsible Team | Estimated Hours | Priority | Dependencies |
| Requirements gathering | Product | 16 | High | None |
| Wireframe design | Design | 24 | High | Requirements gathering |
| UI mockups | Design | 40 | High | Wireframe design |
| Frontend development | Engineering | 80 | High | UI mockups |
Generate a reference table dynamically
Creates a reference table of common HTTP status codes for developer documentation.
=AI_TABLE("HTTP status codes with code number, name, and when it occurs", 15, 3)Output
| Code | Name | When It Occurs |
| 200 | OK | Request succeeded and response contains data |
| 301 | Moved Permanently | Resource has been permanently relocated |
| 404 | Not Found | Requested resource does not exist |
Use Cases
Competitive Landscape Mapping
Quickly generate feature comparison matrices of competing products to inform product strategy and roadmap discussions, then refine with verified data from actual research.
Training Material Creation
Generate reference tables, vocabulary lists, quiz questions, and structured study guides for courses. Teachers can create practice datasets for data analysis classes instantly.
Sample Data for Dashboard Prototyping
Generate realistic-looking datasets to prototype dashboards and reports before production data is available, letting you validate chart formats and KPI structures early in the project.
Sales Battle Cards
Create quick-reference comparison tables showing how your product stacks up against specific competitors on features, pricing, and customer segments for sales team training.
Event Planning Checklists
Generate structured event planning tables with tasks, timelines, responsible parties, and budget allocations for conferences, weddings, product launches, or corporate events.
Regulatory Compliance Matrices
Create comparison tables of regulatory requirements across different markets or jurisdictions, helping compliance teams quickly identify where requirements differ.
Pro Tips
Specify exact column names in your prompt rather than letting the AI choose. This gives you predictable headers that work with VLOOKUP, MATCH, and other Sheets functions that reference column names.
Use AI_TABLE to create template structures, then overwrite individual cells with real data as it becomes available. This gives you a complete-looking spreadsheet immediately while you gather verified information.
Set rows and cols to match your available space. If you have exactly 8 rows of space below the formula, set rows=8 to prevent the output from overlapping other data.
For sample datasets, add realism cues to your prompt: "realistic e-commerce data with varied product categories, different price ranges between $10-$500, and dates throughout January 2025". The more context you give, the more useful the sample data.
This function is a massive time-saver for anyone who regularly creates reference tables, comparison matrices, or sample datasets in spreadsheets. Product managers can generate feature comparison tables in seconds. Teachers can create quiz answer keys or reference charts. Marketers can build competitive landscape tables without hours of manual research. The AI draws on its extensive training data to produce factually grounded tables, though you should verify critical data points for accuracy since AI-generated content can occasionally contain errors.
The optional rows and cols parameters let you control the exact dimensions of the output. If you need a quick 3-column, 5-row table for a presentation, specify those dimensions. If you omit them, the AI decides the appropriate size based on your prompt. AI_TABLE works especially well for brainstorming and prototyping — quickly generate a table structure, review the output, then refine or replace specific cells with verified data. It is also useful for generating realistic sample datasets for testing formulas, pivot tables, and charts before your real data is ready.
Common Errors
#ERROR! Output exceeds cell limitCause: The requested table is too large for Google Sheets to handle in a single array output.
Fix: Reduce the rows or cols parameter. Try generating a smaller table first. For very large tables, split the generation into multiple calls.
#REF! Array output blockedCause: The generated table is trying to spill into cells that already contain data.
Fix: Clear the cells below and to the right of the formula to make room for the table output. The table needs (rows+1) rows and (cols) columns of empty space.
#ERROR! Prompt is too vagueCause: The prompt does not provide enough detail for the AI to generate a meaningful table.
Fix: Be specific about what data you want. Include column names, the subject matter, and any constraints. Example: "Top 10 US cities by population with city, state, population, and area in square miles".
Frequently Asked Questions
AI_TABLE generates data based on the AI model's training knowledge. For factual reference data (like programming languages or HTTP status codes), it is generally accurate. For pricing, market data, or statistics, treat the output as a strong starting point and verify critical figures. The function is especially reliable for structural and comparative information.
Yes, specify column names directly in your prompt. For example: "...with columns: Name, Price, Rating" will produce exactly those three columns. If you also set the cols parameter, make sure the number matches the columns you described in your prompt.
Practically, tables up to about 50 rows and 10 columns work well. Larger tables may hit token limits or take longer to generate. For very large datasets, generate in batches: create the first 50 rows, then use a follow-up prompt for the next 50.
AI_TABLE returns all values as text strings. If you need to perform calculations on generated numbers, wrap them with VALUE() or use 0+cell to convert text to numbers. For currency values, you may need to remove the "$" symbol first using SUBSTITUTE.
AI_TABLE is specifically optimized for tabular output with dimension control (rows and cols parameters). UNLIMITED_AI can also produce tables but gives you less control over exact dimensions. Use AI_TABLE when you need a specific table size and AI for table generation specifically.
Yes, specify the language in your prompt. For example: "Generate a table of 10 Spanish vocabulary words with columns: Spanish, English, Part of Speech". The AI will generate content in whatever language you request.
Related Functions
Start using AI_TABLE today
Install Unlimited Sheets to get AI_TABLE and 41 other powerful functions in Google Sheets.
