G2

Shivansh M.

Consultant Quantum Engineer

Small-Business (50 or fewer emp.)

2/19/2026

Guest users or non-business users of the software, not included in G2 scores.

"Powerful GPU Accelerators and Generous Quotas That Save Time"

5/5

What do you like best about Kaggle?

The hardware accelerators it offers, along with the generous weekly quotas, are a real lifesaver. I often work with the P100 GPU it offers and it has saved so much of my time.

Separating and merging code cells is quite easy with the intuitive keyboard shortcuts. The dual GPU (T4 *2) setup is especially useful for parallel learning. Review collected by and hosted on G2.com.

What do you dislike about Kaggle?

It’s not easy to set up or freeze an environment with specific dependencies. Instead, I have to install the required packages again in every session (or create the dependencies installation file and run it). Also, when I run a cell using the keyboard shortcut (SHIFT + ENTER), it doesn’t visually move me to the next cell, so I end up having to scroll to find it.

Furthermore, it would be nice to be able to see the time it took for each cell to run (similar to google colab). Review collected by and hosted on G2.com.

What problems is Kaggle solving and how is that benefiting you?

Kaggle has become my go-to IDE. Whether it’s a simple test script or a full hybrid Quantum-ML pipeline, I can run it with accelerators that feel better suited to the job on Kaggle (even on the free tier). It took me just under 30 minutes to deploy and run a hybrid QML pipeline using the P100 GPU offered by Kaggle, compared with 3+ hours on just a Colab CPU or 1.5+ hours on a single Colab T4 GPU.

The learning modules also do a good job of teaching the basics of ML and data science, and they provide a plethora of datasets and models to experiment with. Review collected by and hosted on G2.com.