Gemini Now Supports File Uploads & Here’s How It Stacks Up Against ChatGPT

The paid version of Gemini – Gemini Advanced – now lets you upload files to perform data analysis. You can upload several files on Gemini Advanced and ask questions from the documents. The best part is that you can throw multiple files of different formats at Gemini in one go and ask questions related to them. So, I tested Gemini’s file upload feature and compared it with ChatGPT. Here are my findings.

What File Types are Supported on Gemini Advanced?

You can upload different file types such as PDF, CSV, DOCX, XLS, XLSX, TXT, DOTX, RTF, and TSV on Gemini. This basically means you can upload PDF files, Office files like Word documents, Excel spreadsheets, CSV files, and more. There is also support for JPG, JPEG, PNG, and WEBP image files, but these are also available on the free version of Gemini.

Moreover, you can connect your Google Drive via Gemini Extensions and upload Sheets and Docs too.

In comparison to ChatGPT though, the file support is limited on Gemini. ChatGPT can process any kind of file you feed to the chatbot. From ZIP files to coding languages such as JSON, PY, XML; data files like SQL and MATLAB; audio files like WAV, and several other file formats are supported by ChatGPT.

Furthermore, OpenAI offers file uploads to every ChatGPT user for free, but with a capped message limit. In contrast, you will need a Gemini Advanced subscription to be able to upload files, which is surely a downside. Nevertheless, let’s go through the file processing capabilities of Gemini Advanced and see how well it does against ChatGPT.

How to Upload Files on Gemini Advanced

• Go ahead and open gemini.google.com (visit) in your browser.
• Make sure you have subscribed to Gemini Advanced.
• Click on the “+” button at the bottom and then click on “Upload files”.

• After that, upload your local file. I uploaded a large CSV file containing data on the number of people using the Internet around the world from 1990 to 2020 (source).
• Next, I asked a simple question from the uploaded CSV file, and Gemini did a fabulous job despite the file having more than 6,000 rows. ChatGPT also got the answer right on asking the same question. Both models used Python internally to perform data analysis.

• Next, I asked both models to create a chart, and well, it was spot on. Both Gemini Advanced and ChatGPT delivered the same result.

Finally, I uploaded an XLSX spreadsheet containing data on Chromebooks and their specs. I asked both models to find the best Chromebook based on their specifications. Gemini recommended the Samsung Galaxy Chromebook due to its superior display, and ChatGPT recommended the Lenovo Flex 5 because it offers better value for money. I am impressed by both the answers.

Overall, for data processing, both Gemini Advanced and ChatGPT are excellent, powered by Gemini 1.5 Pro and GPT-4o models, respectively. I didn’t see any sign of hallucination, most likely because both models are using Python to parse the data. Google indeed improved its Gemini 1.5 Pro model after Google I/O, and it’s very consistent now.

The only downside with Gemini Advanced is that file upload is not available to free users and file compatibility is also limited. In comparison, ChatGPT offers file uploads to all users, albeit with a limit, and there is support for a wide number of file formats.

It’s down to your preference as to which service you want to choose. Let us know your go-to option in the comments below.