Skill · AI Hiring Index · as of 2026-11-19
PyTorch in AI job postings
8% of technical postings at the AI companies we track mention PyTorch (388 postings). Research Scientist postings ask for it most: 40%.
Technical postings that mention it
8%
388 postings
Change since the previous Index
+1 pt
Which roles ask for it
Share of each technical role family's open postings that mention PyTorch.
| Role family | Share, as a bar | Share |
|---|---|---|
| Research Scientist | 40% | |
| Research Engineer | 35% | |
| Inference / Perf | 32% | |
| AI / ML Engineer | 25% | |
| Evals / Data | 5% | |
| FDE / Applied | 4% | |
| Software Eng | 3% | |
| Infra / Hardware | 3% | |
| Safety / Policy | 2% | |
| Security | 0% |
Companies asking for it most
| Company | Postings | Of its technical postings |
|---|---|---|
| Mistral AI | 34 | 24% |
| Applied Intuition | 22 | 9% |
| Databricks | 21 | 4% |
| Nebius | 20 | 9% |
| Cerebras | 18 | 17% |
| Thinking Machines Lab | 18 | 43% |
| CoreWeave | 17 | 10% |
| Wayve | 17 | 15% |
| Anthropic | 15 | 4% |
| Cohere | 13 | 20% |
How to learn it
- Neural Networks: Zero to Hero — Andrej Karpathy free
Backprop to a GPT in PyTorch, one video at a time — the fastest way for an engineer to read and debug model code.
- Learn the Basics — PyTorch tutorials free
Tensors, autograd, datasets and the training loop in the framework's own words; a reference you will return to.
Project: Train a small GPT and change one thing — Reproduce a small model with nanochat or nanoGPT on a modest dataset, change one design choice, and show the ablation with its loss curves.