Arao AI
Back to search results

gemma-4-12B-it · IQ4_XS

Start a discussion
Apple M2 Ultrallama.cpp v0.42L2 Cross-framework verified

This page aggregates 1 real-world runs of gemma-4-12B-it (IQ4_XS) on Apple M2 Ultra with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

90.66 tok/s

Decode

Decode speed

438.77 tok/s

Prefill

Prefill speed

0.27 s

TTFT

Time to first token

20.75 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 · IQ4_XS (current)Decode 90.66 · Prefill 438.77 ·
  2. llama.cpp · Q5_K_S Decode 66.9 · Prefill 326.75 ·
  3. Ollama · Q2_K Decode 96.52 · Prefill 478.87 ·

Core figures

Decode (avg)
90.66 tok/s
Prefill (avg)
438.77 tok/s
TTFT (avg)
0.27 s
VRAM (avg)
20.75 GB
MTP acceptance rate
—
TTFB
0.73 GB
Power draw
134.7 W

Configuration

Member-level fields are taken from the most recent run

Model
gemma-4-12B-it
Quantization
IQ4_XS
Framework
llama.cpp
Version
v0.42
Context length
2048 tokens
Batch size
256
GPU layers
99
Flash Attention
Off

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)
20.75 GB
OS
macOS 26.1
Power draw
134.7 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

    v0.42 · macOS 26.1 · 2048 ctx

    90.66 tok/s

    Decode

    438.77 tok/s

    Prefill

    0.27 s

    TTFT

    20.75 GB

    VRAM

    —

    MTP

    134.7 W

    Power draw

    L2 batch seed 2026-10-19: llama.cpp on hw_id=103 running model_id=65 at IQ4_XS