Llama-3.2-1B-Instruct

Llama family · 1.0B params

1recorded run

699.9best tok/s

1verified

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

What the Llama-3.2-1B-Instruct numbers mean

The fastest verified result on our board is 699.9 output tok/s on an RTX PRO 6000 Blackwell at Q8_0 with a 4096-token context. That is about 1.4 ms per generated token — 500 tokens takes roughly 0.7 seconds. Prompt processing on the same record runs at 64851.3 tok/s, so a 2,000-token prompt is read in about 0.0 seconds. Peak VRAM was measured at 2.1 GB.

How much VRAM does Llama-3.2-1B-Instruct need?

Llama-3.2-1B-Instruct has 1.0B 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)	1.4 GB	8 GB
Q8_0 (8-bit)	2.0 GB	8 GB
FP16 (16-bit)	3.1 GB	8 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 PRO 6000 Blackwell699.9

Models with a similar footprint

Model	Params	Q4_K_M weights
gemma-3-1b-it	1B	0.6 GB
LFM2.5-1.2B-Instruct	1B	0.6 GB
MiniCPM5-1B	1B	0.6 GB
Qwen3-0.6B	1B	0.6 GB
Qwen3.5-0.8B	800M	0.5 GB
Qwen3.5-0.8B-MTP	800M	0.5 GB

Ranked by distance from Llama-3.2-1B-Instruct's parameter count. Full benchmark board.

All benchmarks Methodology

GPU	Model	Quant	Backend	Ctx	tok/s	prompt tok/s	VRAM GB	Source
RTX PRO 6000 Blackwell · 96 GB	Llama-3.2-1B-Instruct	Q8_0	llama.cpp 876a432	4096	699.9 ✓	64851.3	2.1	src