How much VRAM does Nemotron-Labs-Audex-30B-A3B need?

Q4_K_M is exactly 24,515,130,336 bytes (22.83 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

nvidia · text

Nemotron-Labs-Audex-30B-A3B

nvidia/Nemotron-Labs-Audex-30B-A3B

Nemotron-Labs-Audex-30B-A3B at Q4_K_M is exactly 24,515,130,336 bytes (22.83 GiB / 24.52 GB) — an effective 6.133 bits per weight, not the nominal 4.

From the file· summed from 1 file(s)From the file· KV per layer

Parameters

32.0B

Architecture

nemotron_h_moe

null layers

Context

—

native (config.json)

License

other

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q4_K_M	22.83 GiB	24,515,130,336	6.133	—	LordNeel
Q4_K_L	22.95 GiB	24,647,250,912	6.166	—	LordNeel
Q4_K_M	23.16 GiB	24,866,623,968	6.221	—	LordNeel
Q4_K_L	23.35 GiB	25,073,578,464	6.273	—	LordNeel
Q5_K_M	24.25 GiB	26,040,161,760	6.515	—	LordNeel
Q5_K_L	24.35 GiB	26,150,262,240	6.542	—	LordNeel
Q5_K_M	24.59 GiB	26,404,127,712	6.606	—	LordNeel
Q5_K_L	24.75 GiB	26,576,589,792	6.649	—	LordNeel
Q6_K	31.21 GiB	33,508,167,648	8.383	—	LordNeel
Q6_K_L	31.21 GiB	33,508,167,648	8.383	—	LordNeel
Q8_0	31.28 GiB	33,585,495,712	8.402	—	LordNeel
Q6_K	31.60 GiB	33,934,495,200	8.490	—	LordNeel
Q8_0	31.68 GiB	34,011,823,296	8.509	—	LordNeel

Compare with

same modality, comparable size

vs Qwen3-Coder-30B-A3B-Instruct vs Qwen3.6-27B vs Qwen3.8-27B vs gemma-4-E4B-it vs Qwen3-30B-A3B-Thinking-2507 vs Qwen3-8B All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_K_M at roughly 16.75 GiB. The real file is 22.83 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

from config.json

Layers

—

Attention heads

—

KV heads

—

Head dim

—

Hidden size

—

Vocab

—

Sliding window

none

SWA period

—

MLA

Experts

—

Experts per token

—

use_sliding_window

—

Questions people ask

How much VRAM does Nemotron-Labs-Audex-30B-A3B need?: Q4_K_M is exactly 24,515,130,336 bytes (22.83 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of Nemotron-Labs-Audex-30B-A3B should I use?: Q4_K_M is the usual default. Pick the largest quantization that fits your card at the context you actually need — the table above gives exact sizes for every one published.