Model techniques map
Techniquesinference & servinginference quantization

general family · filed under inference & serving

NVFP4 quantization

Uses NVIDIA’s 4-bit floating-point format for model weights; the evidence includes checkpoints and mixed-precision deployments but does not define one common recipe.

Also called NVFP4 4-bit floating-point quantization, NVFP4 checkpoint, NVFP4 mixed-precision quantization, NVFP4, NVFP4 (NVIDIA 4-bit floating point), NVFP4 quantization checkpoint.

sources
5
models
5
labs adopt it
4
strongest
default

How sources treat it

One count per evidence span, weakest treatment to strongest.

optional 2used 2default 1

Documented in

Further reading

Picked by hand, not extracted: where to read more, not evidence for anything on this page.

Evidence

5 spans quoted from the sources, strongest treatment first.

Only the routed experts are quantized to NVFP4; shared experts and the qkvr linears remain BF16.

defaultpost trainingin InklingThinking Machines Lab

Released under Apache 2.0 with BF16, FP8, NVFP4, and GGUF weights on Hugging Face

usedsoftware implementationin Step 3.7 FlashStepFun

We recommend using the NVFP4 checkpoint nvidia/Qwen3.5-397B-A17B-NVFP4 for optimal serving efficiency on GB200 nodes.

usedinference servingin nvidia/Qwen3.5-397B-A17B-NVFP4NVIDIA

Nemotron 3.5 Lightning ships with an NVFP4 checkpoint alongside BF16, using the same specialized NVFP4 kernels that power Nemotron 3 Ultra across NVIDIA Blackwell, NVIDIA Hopper, and NVIDIA Ampere GPUs.

optionalinference servingin Nemotron 3.5 LightningNVIDIA

NVIDIA's ModelOpt NVFP4 checkpoint

optionalunclearin nvidia/Qwen3.6-35B-A3B-NVFP4NVIDIA

Filed alongside

Other methods under inference & serving :: inference quantization.