Arao AI
Back to search results

gemma-4-31B-it · Q4_K_M

Start a discussion
Apple M2 Ultrallama.cpp v0.27L2 Cross-framework verified

This page aggregates 1 real-world runs of gemma-4-31B-it (Q4_K_M) on Apple M2 Ultra with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

58.99 tok/s

Decode

Decode speed

259.24 tok/s

Prefill

Prefill speed

0.84 s

TTFT

Time to first token

55.48 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

17 days ago

Last verified

Performance

  1. llama.cpp · Q4_K_M (current)Decode 58.99 · Prefill 259.24 ·
  2. Ollama · IQ2_XS Decode 73.22 · Prefill 389.92 ·

Core figures

Decode (avg)
58.99 tok/s
Prefill (avg)
259.24 tok/s
TTFT (avg)
0.84 s
VRAM (avg)
55.48 GB
MTP acceptance rate
—
TTFB
0.47 GB
Power draw
136 W

Configuration

Member-level fields are taken from the most recent run

Model
gemma-4-31B-it
Quantization
Q4_K_M
Framework
llama.cpp
Version
v0.27
Context length
2048 tokens
Batch size
512
GPU layers
99
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)
55.48 GB
OS
macOS 26.1
Power draw
136 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

    v0.27 · macOS 26.1 · 2048 ctx

    58.99 tok/s

    Decode

    259.24 tok/s

    Prefill

    0.84 s

    TTFT

    55.48 GB

    VRAM

    —

    MTP

    136 W

    Power draw

    L2 batch seed 2026-10-19: llama.cpp on hw_id=103 running model_id=63 at Q4_K_M