Data as of Aug 16, 2026 · Based on 3,131,739 AI responses across 10,525 prompts · See how Parse measures this
AWQgithub.comAWQ is an activation-aware weight quantization method for compressing and accelerating large language models (LLMs) to INT3/INT4 precision. It provides efficient CUDA kernels and a TinyChat interface for running LLMs on resource-constrained edge devices like NVIDIA Jetson Orin.
Parse Score
Sources
arxiv.org shapes more of what AI says about AWQ than any other source, at 33% of its citations.
github.com · huggingface.co · lafzusa.com · tensorrigs.com
