Free Kaggle GPU and TPU Quota
You get 30 hours per week of NVIDIA TESLA P100/T4 GPUs. Perfect for TensorFlow, PyTorch, and other GPU-accelerated libraries.
You get 20 hours per week of Tensor Processing Units (TPUs). Built specifically for heavy deep learning, allowing up to 9-hour continuous sessions.
How to Claim this Deal
- 1
- 2Verify your account with a phone number (this is mandatory to prevent bot abuse).
- 3Click Create - Notebook button from left side menu.
- 4Go to the settings menu on the right and change the "Accelerator" from None to GPU or TPU.
- 5Manually stop your interactive sessions when you are done coding so you don't burn your weekly hours on idle time limit.
Key Benefits
- Free access to NVIDIA P100/T4 GPUs
- Free access to Google TPUs
- Up to 9-hour continuous session times
- Zero local hardware required
- Pre-configured ML environments out of the box
Eligibility & Requirements
| Requirements | • No credit card required |
|---|---|
| Criteria |
|
Kaggle (owned by Google) is the internet's hub for data science and machine learning. Besides hosting massive datasets and coding competitions, they provide a free Jupyter notebook environment backed by Google's servers. It's the go-to place for developers to build, train, and test models without having to buy expensive local hardware.
