Introduction

42 Rows is an advanced AI-powered spreadsheet tool designed to automate complex processes and amplify the capabilities of existing AI models. Our platform allows you to run thousands of prompts automatically based on your template, enabling AI to execute long and intricate tasks that would otherwise be time-consuming.

With 42 Rows, users can either upload a CSV file or create a table directly on our website. Once the table is set up, you can write a prompt and choose an AI model to generate text, create images, or analyze images. The tool then executes these tasks for every row in the table, streamlining your workflow and saving valuable time.

Key Features

• Generate and Manipulate Complex Spreadsheets with AI Technologies
• Extract Key Insights from Websites
• Generate text, images, or analyze images automatically
• Create tailor-made AI solutions

How to create a table

Click on the “File” tab in the menu, then hit the “New” button. Enter the table name (“Table Name“) and the number of columns you need (“Adding Column Count“).

This action will create a table with one row and the specified number of columns.

How to import a table

Click on the “File” tab in the menu, then hit the “Load CSV” button. Select your CSV or JSON file from your computer.

The data in the file will immediately appear in a table on the page.

How to open an existing table

If you’ve previously edited a table and want to open it again, click on the “File” tab in the menu, then hit the “Open” button. Choose the existing file you want to open.

The file will be immediately displayed in a table on the page.

How to save a table in CSV format

Click on the “File” tab in the menu, then hit the “Save” button – then “To CSV“.

The file will be immediately downloaded.

How to add a table to current training

Click on the “File” tab in the menu, then hit the “Add Current Table to Embeddings” button.

The file will be immediately sent to AI and the information will be available to be used in the next operations.

How to add columns or rows to the table

Click on the “Table” tab in the menu, then hit the “Add” button. It will shows two buttons:

– “Rows“: to add rows to the table

– “Columns“: to add columns to the table

In the pop-up insert the number of rows/columns you want to add.

The rows/columns will be automatically added to your table.

How to clean/remove rows or columns in the table

Click on the “Table” tab in the menu, then hit the “Remove” button. It will shows four buttons:

– “Clean Row“: to clear all the data in a selected row

– “Clean Column“: to clear all the data in a selected column

– “Remove Row“: to remove a selected row from the table

– “Remove Column“: to remove a selected column from the table

The rows/columns will be automatically cleaned/removed from your table.

How to merge different columns in the table

Click on the “Table” tab in the menu, then hit the “Merge Column” button.

Select the columns you want to merge.

The columns will automatically merge in your table.

How to copy and add a column from another file in a table

Click on the “Table” tab in the menu, then hit the “Copy Table Data” button.

In the pop-up, select the column where you want to overwrite the data through “Current Table Columns” , choose the file (“Choose a file“) where to find the data and the column you want to copy (“Selected Table Columns“)

The column will be automatically copied in your table, in the selected column.

How to send a column/row to AI

Click on the “Table” tab in the menu, then hit the “Send to Row” or “Send to Column” button.

In the pop-up, select the AI model you want to use (first box), if you want to overwrite or edit the data (second box) and the prompt to tell the AI the action you want it to perform.

The row/column will be automatically sent to the AI and processed with your prompt.

How to invert columns and rows

Click on the “Table” tab in the menu, then hit the “Invert Column and Row” button.

The rows/columns will be automatically inverted.

How to change order of columns and rows

Click on the “Table” tab in the menu, then hit the “Change order” button.

In the pop-up, drag the column you want to move in the new order.

The columns will be re-ordered on the table.

How to set columns prompts

Click on the column settings (three dots on the column header), then hit the “Set Prompt” button.

In the pop-up that opens, perform the following actions:

1) Select the type of input you want to send to the model:

– Prompt – to input text

– Image – to input an image

– Template – to input a template

– Vision – to input a vision

2) Enter a column title in the “Title” box

3) Choose the AI model that will process the prompt in the “Select model” box

4) Select the column you want to reference to in the prompt in the “Select the column to set Prompts” box

5) Write the prompt you want to send to the AI model selected in the “Prompt” box

6) Specify whether the AI should generate content based on information already present in the table.

Specifically, generate content based on:

– None – no other information

– Row – information in the rows

– Table – information across the entire table

– Column – information in the columns

Then, save the prompt and close the pop-up.

How to send prompts to AI

In 42rows, you can send prompts in different ways:

1) Send column – send the prompt of a specific column to the AI

Open the column settings (three dots on the column header) and click the button “Send Column“.

The column’s prompt will be processed and the result will be showed on the specific column.

2) Send selected rows – send the prompt of a specific cohort of rows to the AI

Select the rows you want to send to the AI through the lateral flag and click the button “Send Selected”.

All the prompts will be processed and the results will be shown for the specific rows.

3) Send all – send all the prompts of the table to the AI.

Click the button “Send All“.

All the prompts will be processed and the results will be shown on the table.

It is possible to overwrite the existing fields through the flag “Overwrite filled fields“.

Every time you click a “Send” button, an alert will appear previewing the cost of the operations (in term of credits).

How to copy the prompts from another table

Click on the “Table” tab in the menu, then hit the “Prompts” button.

Click on “Copy“, and select the table you want to copy the prompts from.

Click on “Save” and the prompts will be copied on your table.

The Chatbot feature provides users with a versatile and powerful interface to interact with various AI language models. It offers multiple functionalities, including model selection, response type selection, and data export options.

How to manage a table with Chatbot

Click on the “Chat Bot” button and select an AI model.

Flag the button “Send Table Data” if you want to use the existing table as input.

Select the response type:

– GPT Response: The AI will provide a standard conversational response, similar to ChatGPT.

– Embeddings Response: The AI will use embeddings to generate an answer, which can provide more context-aware responses.

– Table Response: The AI will generate a response based on information from a provided table.

Write your prompt in the text box.

The AI will generate a table as response.

Hit the “Get CSV” button to open the new table in 42rows, otherwise “Copy” or “Cancel” the table through the buttons.

Layouts are pre-made templates that can be used to quickly set a table for a specific matter (e.g. Shopify CSV catalogue).

How to set a file layout

Click on the “Layout” tab in the menu, then hit the “File Layouts” button.

Select the file layout you want to use from the list that opens.

These layout are currently available:

– Google Merchant feed

– Shopify products

– Prestashop categories

– Prestashop combinations

– Magento products

– Amazon.it inventory

– OpenCart product

The table will automatically open in 42rows.

 

This modal window allows users to enhance 42 Rows’ content generation capabilities by providing custom training data.

There are two different ways to train the ChatBot:

1) Chatbot training from the link

Input fields for pasting URLs (multiple URLs allowed, separated by commas or newlines) in the dedicated box.

Click “Train you chatbot” and the input will be added as training.

2) Chatbot training from the file

Input fields by selecting your own files from your PC (supported formats: PDF, TXT; Batch upload enabled).

Click “Train your chatbot” and the input will be added as training.

It is possible to manage all the training from the “Manage your ChatBot” box.

The following actions are available:

– “Test your ChatBot” to try the new trained ChatBot with a new prompt

– “Show ChatBot trainings” to view a list of all added traning

– “Delete all trainings” to delete all added training

The scraping function allows users to collect and organize information from a website into a CSV table.

 

How to do scraping with 42rows

Click on the “Scraping” tab in the menu, then select the type of scraping you want the AI to perform.

Insert the training page link in the “URL” box.

Insert the limit of elements you want to add to your table.

Flag “Add Data in Training” to train the AI with the file linked before.

Flag “Get Product Information” if you are scraping an e-commerce and you need to store in the table the product information.

Flag “Create Table From Result” to generate a table and directly open it in 42rows.

AI models available

Users can choose from a wide range of AI language models:

– GEMINI-1.5-PRO

– GEMINI-1.5-FLASH

– GPT-3.5

– GPT-4

– GPT-4O

– GPT-4O-MINI

– O1-PREVIEW

– O1-MINI

– GPT-4-1106-preview

– GPT-3.5-turbo-1106

– GPT-3.5-turbo-16k

– GPT-3.5-turbo-instruct

– Anthropic

These models are available in a single dropdown selector, allowing users to switch between different AI capabilities easily.

 

How to select the best AI model

Choosing an AI model could be difficult without a clear view of their features. In the following lines, will be listed all the advantages for each model.

– GEMINI-1.5-PRO

Advantages:

1. High precision and language fluency, with advanced problem-solving capabilities.

2. Can handle complex reasoning tasks, making it versatile across industries.

3. Optimized for professional and enterprise-level applications.

– GEMINI-1.5-FLASH

Advantages:

1. Fast processing speeds with slightly reduced depth in reasoning.

2. Optimized for real-time interactions.

3. Lower cost compared to GEMINI-1.5-PRO.

– GPT-3.5

Advantages:

1. Versatile, powerful, and widely used for general-purpose tasks.

2. Lower cost compared to GPT-4 while maintaining strong performance for standard tasks.

– GPT-4

Advantages:

1. Superior accuracy, deep reasoning capabilities, and nuanced understanding.

2. Handles more complex prompts and generates higher quality output than GPT-3.5.

– GPT-4O-MINI

Advantages:

1. A lighter, faster variant of GPT-4O, offering a balance between performance and cost.

2. Efficient for standard operations without needing the full depth of GPT-4O.

– GPT-4O

Advantages:

1. Optimized version of GPT-4 for more specialized tasks, potentially with improved efficiency for specific use cases.

2. Balances performance and cost with focus on optimization.

– O1-PREVIEW

Advantages:

1. Cutting-edge model currently in preview, designed for a wide range of applications.

2. Offers flexibility in experimentation with the latest advancements in AI.

– O1-MINI

1. Advantages:

2. Lightweight and efficient version of O1, optimized for lower resource use while maintaining reasonable accuracy.

– GPT-4-1106-preview

Advantages:

1. Preview version of GPT-4 with some additional features or optimizations.

2. Ideal for users wanting to explore the latest enhancements in GPT-4.

– GPT-3.5-turbo-1106

Advantages:

1. Optimized variant of GPT-3.5, offering faster responses and lower costs while maintaining quality.

– GPT-3.5-turbo-16k

Advantages:

1. Enhanced GPT-3.5 model capable of handling longer inputs and outputs (16k tokens).

2. Efficient for tasks requiring larger datasets.

– GPT-3.5-turbo-instruct

Advantages:

1. Specifically designed to follow instructions more precisely compared to standard GPT-3.5 models.

– Anthropic

Advantages:

1. Focused on providing more ethical and explainable AI outputs.

2. Trained with safety and alignment in mind, suitable for sensitive environments.

Notes

• The availability and performance of different AI models may vary.
• The Embeddings and Table response types may provide more specialized or data-driven answers compared to the standard GPT response.
• Always review the exported data for accuracy and relevance to your needs.

This chatbot system offers a flexible and powerful way to interact with various AI models, making it suitable for a wide range of applications, from casual conversation to data-driven analysis.