Llama-3.2-3B-Instruct

Llama family · 3.0B params

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

What the Llama-3.2-3B-Instruct numbers mean

The fastest verified result on our board is 489.2 output tok/s on an RTX PRO 6000 Blackwell at Q4_K_M with a 640-token context. That is about 2.0 ms per generated token — 500 tokens takes roughly 1.0 seconds. Prompt processing on the same record runs at 28478.5 tok/s, so a 2,000-token prompt is read in about 0.1 seconds. Peak VRAM was measured at 2.6 GB.

How much VRAM does Llama-3.2-3B-Instruct need?

Llama-3.2-3B-Instruct has 3.0B 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.

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.

Models with a similar footprint

Ranked by distance from Llama-3.2-3B-Instruct's parameter count. Full benchmark board.

All benchmarks Methodology