NVIDIA-Nemotron-Nano-12B-v2

Nemotron family · 12.0B params

1recorded run

40.4best tok/s

1verified

Evidence: 1 sourced run · 1 independent source · first recorded Jul 12, 2026. What these labels mean.

What the NVIDIA-Nemotron-Nano-12B-v2 numbers mean

The fastest verified result on our board is 40.4 output tok/s on an RTX 3060 x2 at Q4_K_M with a 512-token context. That is about 24.8 ms per generated token — 500 tokens takes roughly 12.4 seconds. Prompt processing on the same record runs at 1016.0 tok/s, so a 2,000-token prompt is read in about 2.0 seconds. Peak VRAM was measured at 9.9 GB.

How much VRAM does NVIDIA-Nemotron-Nano-12B-v2 need?

NVIDIA-Nemotron-Nano-12B-v2 has 12.0B parameters. These estimates use the same formula as the VRAM calculator: weights, plus about 15% for a moderate-context KV cache and runtime overhead, plus a small engine floor.

Quantisation	Estimated VRAM	Smallest card that fits
Q4_K_M (4-bit)	8.6 GB	12 GB
Q8_0 (8-bit)	15.5 GB	24 GB
FP16 (16-bit)	28.4 GB	48 GB

Long contexts and large batches raise the real number — read the smallest card as "possible with a reduced context". Measured throughput is on the benchmark board.

Fastest GPUs for this model (best tok/s)

RTX 3060 x240.4

Models with a similar footprint

Model	Params	Q4_K_M weights
gemma-3-12b-it	12B	6.8 GB
gemma-4-12B-coder-fable5-composer2.5-v1	12B	6.8 GB
gemma-4-12B-it	12B	6.8 GB
gemma-4-12B-it-qat	12B	6.8 GB
GLM-4.6V-Flash	10B	5.6 GB
InternVL3-14B	14B	7.9 GB

Ranked by distance from NVIDIA-Nemotron-Nano-12B-v2's parameter count. Full benchmark board.

All benchmarks Methodology

GPU	Model	Quant	Backend	Ctx	tok/s	prompt tok/s	VRAM GB	Source
RTX 3060 x2 · 12 GB	NVIDIA-Nemotron-Nano-12B-v2	Q4_K_M	llama.cpp 13f2b28	512	40.4 ✓	1016.0	9.9	src