deployedbyai

The Deploy Log | Models and APIs

CUDA kernel agent skill

Hugging Face released an agent skill that teaches Codex and Claude to write production CUDA kernels, producing an RMSNorm kernel for Qwen3-8B with an average 1.94x speedup on H100.

Hugging Face | In Edition 21, Saturday 14 Feb 2026

SOURCE Hugging Face, Custom Kernels for All from Codex and Claude

Added to The Deploy Log Tuesday 15 Sep 2026, updated Thursday 17 Sep 2026. Read Edition 21, the edition that carried it.

The call on this one

What shipped is free. The call on it, what to do about it and the condition on that, opens with a signup: free, no card, every edition in full.

It becomes $50 a year, and signing up now keeps your first year free.