Arao AI
Back to search results

Qwen2.5-Omni-3B · Q6_K

Start a discussion
Apple M5 Max128GBllama.cpp v0.29L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen2.5-Omni-3B (Q6_K) on Apple M5 Max with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

221.16 tok/s

Decode

Decode speed

1,117.52 tok/s

Prefill

Prefill speed

0.72 s

TTFT

Time to first token

8.4 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

2 days ago

Last verified

Performance

  1. llama.cpp · IQ3_S Decode 341.31 · Prefill 1584.99 ·
  2. llama.cpp · Q6_K (current)Decode 221.16 · Prefill 1117.52 ·
  3. Strata · Q8_0 Decode 241.28 · Prefill 991.2 ·
  4. llama.cpp · IQ1_S Decode 289.86 · Prefill 1338.5 ·

Core figures

Decode (avg)
221.16 tok/s
Prefill (avg)
1,117.52 tok/s
TTFT (avg)
0.72 s
VRAM (avg)
8.4 GB
MTP acceptance rate
—
TTFB
1.77 GB
Power draw
274.2 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen2.5-Omni-3B
Quantization
Q6_K
Framework
llama.cpp
Version
v0.29
Context length
32768 tokens
Batch size
512
GPU layers
20
Flash Attention
Off

Hardware

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

GPU
Apple M5 Max
Nominal VRAM
128 GB
Measured VRAM (avg)
8.4 GB
OS
Ubuntu 24.04
Driver
565.57
CUDA
12.4
Power draw
274.2 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-16

    v0.29 · Ubuntu 24.04 · CUDA 12.4 · 32768 ctx

    221.16 tok/s

    Decode

    1,117.52 tok/s

    Prefill

    0.72 s

    TTFT

    8.4 GB

    VRAM

    —

    MTP

    274.2 W

    Power draw

    L2 batch seed 2026-10-16: llama.cpp on hw_id=28 running model_id=60 at Q6_K