Nanbeige4.3-33B

Nanbeige family · 33.0B params

1recorded run

66.2best tok/s

1verified

Evidence: 1 sourced run · 1 independent source · first recorded Oct 26, 2026. What these labels mean.

What the Nanbeige4.3-33B numbers mean

The fastest verified result on our board is 66.2 output tok/s on an RTX 3090 x2 at W4A16 (INT4 AutoRound, GPTQ, gro with a 131072-token context. That is about 15.1 ms per generated token — 500 tokens takes roughly 7.6 seconds.

How much VRAM does Nanbeige4.3-33B need?

Nanbeige4.3-33B has 33.0B 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)	22.1 GB	32 GB
Q8_0 (8-bit)	41.1 GB	64 GB
FP16 (16-bit)	76.7 GB	96 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 3090 x266.2

Models with a similar footprint

Model	Params	Q4_K_M weights
gemma-4-31B-it	33B	18.6 GB
GLM-Z1-32B-0414	33B	18.6 GB
Laguna-XS-2.1	33B	18.6 GB
Nemotron-Cascade-2-30B-A3B	32B	18.0 GB
NVIDIA-Nemotron-3-Nano-30B-A3B	32B	18.0 GB
NVIDIA-Nemotron-3.5-Lightning-30B-A3B	32B	18.0 GB

Ranked by distance from Nanbeige4.3-33B's parameter count. Full benchmark board.

All benchmarks Methodology

GPU	Model	Quant	Backend	Ctx	tok/s	prompt tok/s	VRAM GB	Source
RTX 3090 x2 · 24 GB	Nanbeige4.3-33B	W4A16 (INT4 AutoRound, GPTQ, gro	vllm v0.27.1	131072	66.2 ✓	—	—	src