Wikitopia
AWQ
conceptQuantization Method
Try in Playground →RSS
Overview
Developed byMIT Han Lab
LicenseMIT License
Open source✓ Open Source
Use case4-bit quantization of large language models
Also see
Alternative to
GPTQSmoothQuant
Knowledge graph stats
Claims18
Avg confidence93%
Avg freshness100%
Last updatedUpdated 2 days ago
Trust distribution
100% unverified
Governance
Not assessed
Contribute governance data →

AWQ

concept

Activation-aware Weight Quantization method for efficient 4-bit quantization of LLMs preserving important weights.

Compare with...

published year

ValueTrustConfidenceFreshnessSources
2023○UnverifiedHighFresh1

full name

ValueTrustConfidenceFreshnessSources
Activation-aware Weight Quantization○UnverifiedHighFresh1

quantization bits

ValueTrustConfidenceFreshnessSources
4-bit○UnverifiedHighFresh1

primary use case

ValueTrustConfidenceFreshnessSources
4-bit quantization of large language models○UnverifiedHighFresh1
Large Language Model Quantization○UnverifiedHighFresh1

developed by

ValueTrustConfidenceFreshnessSources
MIT Han Lab○UnverifiedHighFresh1
MIT○UnverifiedHighFresh1

open source

ValueTrustConfidenceFreshnessSources
true○UnverifiedHighFresh1

implemented by

ValueTrustConfidenceFreshnessSources
AutoAWQ○UnverifiedHighFresh1
TinyChat○UnverifiedModerateFresh1

supports model

ValueTrustConfidenceFreshnessSources
LLaMA○UnverifiedHighFresh1
OPT○UnverifiedHighFresh1

alternative to

ValueTrustConfidenceFreshnessSources
GPTQ○UnverifiedHighFresh1
SmoothQuant○UnverifiedHighFresh1

license type

ValueTrustConfidenceFreshnessSources
MIT License○UnverifiedHighFresh1

integrates with

ValueTrustConfidenceFreshnessSources
PyTorch○UnverifiedHighFresh1
Hugging Face Transformers○UnverifiedModerateFresh1
vLLM○UnverifiedModerateFresh1

Alternatives & Similar Tools

GPTQ
alternative to
Compare →
SmoothQuant
alternative to
Compare →

Commonly Used With

PyTorchHugging Face TransformersvLLM

Related entities

Claim count: 18Last updated: 4/8/2026Edit history