The source code used for this example is available in our GitHub repository.

This chapter demonstrates how to use Cohere and Google Apps Script, a scripting language based on JavaScript, to build two custom functions that add large language model (LLM) capabilities to Google Sheets documents. We'll be generating summaries of movie reviews.

It involves the following steps.

Step 1: Create a custom Apps Script function for text summarization.
Step 2: Run summarization on a new input.

Setup

We’ve already created a file to run the demo of this project. Open the file in Google Sheets and follow the instructions described in the Instructions worksheet.

Once you have completed the steps, you can run the demo directly using the script provided in the repository. Let's see how it is created.

Step 1: Create a Custom Apps Script Function for Text Summarization

Here, we’ll be using the Chat endpoint instead.

We create a simple prompt asking the endpoint to provide a summary in about ten words. We are defining three parameters for the endpoint call.

message - the prompt for the LLM
temperature - we keep the value closer to 0 to make the output more predictable (0.2 in this example)
model- we'll use command-r-plus-08-2024

Finally, we parse the response and output the summary.

function summarize(cell_value) {

    let raw = JSON.stringify({
        "message": "Summarize this in about 10 words: " + cell_value,
        "temperature": 0.2,
        "model": "command-r-plus-08-2024",
    });
    let requestOptions = {
        'method': 'post',
        'muteHttpExceptions': true,
        'contentType': 'application/json',
        'headers': {
            'Authorization': 'Bearer ' + TOKEN
        },
        'payload': raw,
        redirect: 'follow'
    };
    
     let response = UrlFetchApp.fetch("https://api.cohere.ai/v1/chat", requestOptions)
     let responseContentTxt = JSON.parse(response.getContentText());
     let summarizedTxt = "SUMMARY: " + responseContentTxt.text + "\n";
     return summarizedTxt;

}

Finally, we save the code we have added for text summarization.

Step 2: Run Text Summarization on a New Input

Now, we get to the part we’ve been waiting for – classifying and summarizing text on Google Sheets.

For this, we’ll move back to the Google Sheets view and go to the Demo worksheet. It has two columns:

The first column is the input text, the movie review we want to summarize. Here, we have a sample of a few reviews, but you can add as many as you want.
In the second column, we call our text summarization function by specifying =summarize(A3).

This takes cell the movie reviews in the column A and outputs the summary.

Conclusion

Dealing with numbers in a spreadsheet has always been easy, but the same couldn’t be said of text — until now. With Cohere’s natural language processing capabilities, you can supercharge your spreadsheets, saving time and effort.

Deploying on Google Sheets with Google Apps Script