Arao AI
Back to search results

Qwen3.5-9B · Q3_K_S

Start a discussion
Apple M1 ProvLLM v0.47L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-9B (Q3_K_S) on Apple M1 Pro with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

79.35 tok/s

Decode

Decode speed

395.79 tok/s

Prefill

Prefill speed

0.72 s

TTFT

Time to first token

14.38 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_XS Decode 75.72 · Prefill 277.89 ·
  2. vLLM · IQ2_M Decode 84.15 · Prefill 281.3 ·
  3. vLLM · Q3_K_S (current)Decode 79.35 · Prefill 395.79 ·

Core figures

Decode (avg)
79.35 tok/s
Prefill (avg)
395.79 tok/s
TTFT (avg)
0.72 s
VRAM (avg)
14.38 GB
MTP acceptance rate
—
TTFB
0.63 GB
Power draw
39.7 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-9B
Quantization
Q3_K_S
Framework
vLLM
Version
v0.47
Context length
4096 tokens
Batch size
32
GPU layers
999
Flash Attention
On

Hardware

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

GPU
Apple M1 Pro
Measured VRAM (avg)
14.38 GB
OS
macOS 15.0
Driver
Apple M1 Pro
CUDA
N/A
Power draw
39.7 W

Sources & evidence

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

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

    v0.47 · macOS 15.0 · CUDA N/A · 4096 ctx

    79.35 tok/s

    Decode

    395.79 tok/s

    Prefill

    0.72 s

    TTFT

    14.38 GB

    VRAM

    —

    MTP

    39.7 W

    Power draw

    L2 batch seed 2026-10-16: vLLM on hw_id=100 running model_id=49 at Q3_K_S