Arao AI
Back to search results

Llama-3.2-3B-Instruct · Q5_K_M

Start a discussion
Apple M2 UltraOllama v0.21L2 Cross-framework verified

This page aggregates 1 real-world runs of Llama-3.2-3B-Instruct (Q5_K_M) on Apple M2 Ultra with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

207.01 tok/s

Decode

Decode speed

946.05 tok/s

Prefill

Prefill speed

1.6 s

TTFT

Time to first token

7.96 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

16 days ago

Last verified

Performance

  1. llama.cpp · Q5_K_M Decode 233.78 · Prefill 783.42 ·
  2. Ollama · Q5_K_M (current)Decode 207.01 · Prefill 946.05 ·

Core figures

Decode (avg)
207.01 tok/s
Prefill (avg)
946.05 tok/s
TTFT (avg)
1.6 s
VRAM (avg)
7.96 GB
MTP acceptance rate
—
TTFB
1.66 GB
Power draw
122.7 W

Configuration

Member-level fields are taken from the most recent run

Model
Llama-3.2-3B-Instruct
Quantization
Q5_K_M
Framework
Ollama
Version
v0.21
Context length
32768 tokens
Batch size
256
GPU layers
33
Flash Attention
On

Hardware

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

GPU
Apple M2 Ultra
Measured VRAM (avg)
7.96 GB
OS
macOS 26.1
Power draw
122.7 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-19

    v0.21 · macOS 26.1 · 32768 ctx

    207.01 tok/s

    Decode

    946.05 tok/s

    Prefill

    1.6 s

    TTFT

    7.96 GB

    VRAM

    —

    MTP

    122.7 W

    Power draw

    L2 batch seed 2026-10-19: Ollama on hw_id=103 running model_id=70 at Q5_K_M