Arao AI
Back to search results

gemma-4-12B-it · Q2_K

Start a discussion
Apple M4 Pro64GBvLLM v0.28L2 Cross-framework verified

This page aggregates 1 real-world runs of gemma-4-12B-it (Q2_K) on Apple M4 Pro with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

110.03 tok/s

Decode

Decode speed

407.67 tok/s

Prefill

Prefill speed

2.35 s

TTFT

Time to first token

15.96 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

5 days ago

Last verified

Performance

  1. llama.cpp · IQ2_M Decode 103.7 · Prefill 452.45 ·
  2. llama.cpp · Q5_K_S Decode 68.46 · Prefill 358.45 ·
  3. vLLM · Q2_K (current)Decode 110.03 · Prefill 407.67 ·
  4. vLLM · Q4_K_S Decode 90.85 · Prefill 484.54 ·

Core figures

Decode (avg)
110.03 tok/s
Prefill (avg)
407.67 tok/s
TTFT (avg)
2.35 s
VRAM (avg)
15.96 GB
MTP acceptance rate
—
TTFB
0.88 GB
Power draw
67.1 W

Configuration

Member-level fields are taken from the most recent run

Model
gemma-4-12B-it
Quantization
Q2_K
Framework
vLLM
Version
v0.28
Context length
32768 tokens
Batch size
32
GPU layers
999
Flash Attention
Off

Hardware

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

GPU
Apple M4 Pro
Nominal VRAM
64 GB
Measured VRAM (avg)
15.96 GB
OS
macOS 15.0
Driver
Apple M4 Pro
CUDA
N/A
Power draw
67.1 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-13

    v0.28 · macOS 15.0 · CUDA N/A · 32768 ctx

    110.03 tok/s

    Decode

    407.67 tok/s

    Prefill

    2.35 s

    TTFT

    15.96 GB

    VRAM

    —

    MTP

    67.1 W

    Power draw

    L2 batch seed 2026-10-13: vLLM on hw_id=25 running model_id=65 at Q2_K