GeminiWithFiles

This is a Google Apps Script library for Gemini API with files.

AI Summary

GeminiWithFiles

A Google Apps Script wrapper that uploads files to the Gemini API and handles generative-AI interactions

Target Users

Intermediate-to-advanced GAS developers who need to analyse or generate content from Drive-stored images, PDFs, etc. via the Gemini API

Problems Solved

Passing files from GAS to Gemini requires handling MIME conversion, 50 MB limits and async uploads, which is cumbersome to implement manually

Main Features

One-liner upload of Drive files

Simply pass file IDs or Blobs to upload images, PDFs or videos asynchronously; resumable upload handles files larger than 50 MB

Batch generation & multimodal support

Feed Gemini up to thousands of images or many files at once and obtain descriptions, summaries, etc. in a single API call

Output controlled via JSON schema

Specify response_mime_type and schema to receive strictly formatted JSON, simplifying downstream processing

Chat, token counting & utilities

Provides chat history, token counting, function calling and system instructions as straightforward helper methods

Usage Examples

Upload one image and get a description

function quickDemo() {
  const apiKey = 'YOUR_API_KEY';           // Acquire from MakerSuite
  const fileId = 'DRIVE_FILE_ID';          // Image to describe

  // When the library is installed
  const g = GeminiWithFiles.geminiWithFiles({ apiKey });

  // Upload the image
  const files = g.setFileIds([fileId]).uploadFiles();

  // Ask Gemini for a caption
  const res = g.withUploadedFilesByGenerateContent(files)
               .generateContent({ q: 'Describe this image in under 50 words.' });

  Logger.log(res); // => Generated caption
}

This minimal snippet uploads a Drive image and asks Gemini for a short caption. Pass multiple IDs to setFileIds() for batch processing.

GeminiWithFiles

AI Summary