Model techniques map
Techniquesinference & servinginference quantization

general family · filed under inference & serving

Quantization

A broad family of methods that reduce numerical precision to reduce model memory use and potentially accelerate inference.

Also called quantization algorithms, quantized versions.

sources
3
models
2
labs adopt it
2
strongest
core

How sources treat it

One count per evidence span, weakest treatment to strongest.

optional 2core 1

Documented in

Evidence

3 spans quoted from the sources, strongest treatment first.

Quantization is a key technology to run models on consumer hardware by reducing their memory footprint while also accelerating decode speed.

coreoptimizationin Gemma 4Google

quantized versions run natively on consumer GPUs to power your IDEs, coding assistants and agentic workflows

optionalinference servingin Gemma 4Google

AngelSlim supports a comprehensive suite of compression tools ... including common quantization algorithms, low-bit quantization

optionalsoftware implementationin AngelSlimTencent Hunyuan

Filed alongside

Other methods under inference & serving :: inference quantization.