HeyTensor
Fix PyTorch tensor shape errors quickly.
Pricing: free — Free to use in browser · Visit website
HeyTensor is a web-based tool that helps you calculate and debug tensor shapes for various layers in PyTorch models. It supports Conv2d, Linear, LSTM, and more, allowing you to chain layers and explore architecture presets. Simply paste your error message to get instant fixes.
Pros
- Instantly fix tensor shape errors
- Supports multiple layer types
- Works directly in your browser
Cons
- Limited to PyTorch models only
- No advanced customization options
FAQ
How do I calculate Conv2d output shape?
Use the formula: H_out = floor((H_in + 2*padding - dilation*(kernel_size-1) - 1) / stride) + 1.
Why does PyTorch say mat1 and mat2 shapes cannot be multiplied?
Mismatch in the number of columns in mat1 and rows in mat2. Ensure they match for Linear layers.
Is HeyTensor free to use?
Yes, it's free and runs directly in your browser.
Top alternatives
TensorZero offers tools for LLM applications.
TensorZero offers an open-source alternative for developers to build and deploy machine learning models.
Track updates for AI coding tools like Claude Code and Copilot.
Havoptic offers a freemium model with advanced visualizations for developers, complementing HeyTensor's free web-based AI services.
Warp is a modern AI terminal for coding with agents.
Warp offers a user-friendly interface for building and deploying machine learning models, similar to HeyTensor's approach in simplifying AI
AI-driven bug fixing assistant for developers.
Goast AI offers a freemium model with advanced developer tools, complementing HeyTensor's free web-based services.
Open-source ML library for Python and C++.
TensorFlow offers a robust framework for machine learning development, contrasting with HeyTensor's user-friendly web interface.
Last updated: 2026-07-30