Arao AI
Back to search results

Llama-3.2-3B-Instruct · IQ3_S

Start a discussion
NVIDIA RTX 4060 Ti 16GB16GBllama.cpp v0.36L2 Cross-framework verified

This page aggregates 1 real-world runs of Llama-3.2-3B-Instruct (IQ3_S) on NVIDIA RTX 4060 Ti 16GB with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

326.89 tok/s

Decode

Decode speed

1,104.91 tok/s

Prefill

Prefill speed

0.63 s

TTFT

Time to first token

6.26 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · IQ3_S (current)Decode 326.89 · Prefill 1104.91 ·
  2. vLLM · Q2_K Decode 350.23 · Prefill 1317.38 ·

Core figures

Decode (avg)
326.89 tok/s
Prefill (avg)
1,104.91 tok/s
TTFT (avg)
0.63 s
VRAM (avg)
6.26 GB
MTP acceptance rate
—
TTFB
2.62 GB
Power draw
150.9 W

Configuration

Member-level fields are taken from the most recent run

Model
Llama-3.2-3B-Instruct
Quantization
IQ3_S
Framework
llama.cpp
Version
v0.36
Context length
2048 tokens
Batch size
256
GPU layers
20
Flash Attention
On

Hardware

Nominal and measured figures are shown side by side; whether it runs is the reader's call

GPU
NVIDIA RTX 4060 Ti 16GB
Nominal VRAM
16 GB
Measured VRAM (avg)
6.26 GB
OS
Ubuntu 24.04
Driver
572.83
CUDA
12.8
Power draw
150.9 W

Sources & evidence

1 measured records in total, each traceable to its original source

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-18

    v0.36 · Ubuntu 24.04 · CUDA 12.8 · 2048 ctx

    326.89 tok/s

    Decode

    1,104.91 tok/s

    Prefill

    0.63 s

    TTFT

    6.26 GB

    VRAM

    —

    MTP

    150.9 W

    Power draw

    L2 batch seed 2026-10-18: llama.cpp on hw_id=3 running model_id=70 at IQ3_S