Arao AI
Back to search results

Qwen3.5-0.8B · Q4_K_S

Start a discussion
Apple M4 Max128GBllama.cpp v0.41L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-0.8B (Q4_K_S) on Apple M4 Max with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

304.54 tok/s

Decode

Decode speed

1,527.8 tok/s

Prefill

Prefill speed

1.17 s

TTFT

Time to first token

3.41 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · IQ4_XS Decode 313.09 · Prefill 1177.06 ·
  2. llama.cpp · Q4_K_S (current)Decode 304.54 · Prefill 1527.8 ·
  3. Strata · Q2_K Decode 352.85 · Prefill 1792.79 ·
  4. Strata · Q3_K_L Decode 303.63 · Prefill 1479.8 ·

Core figures

Decode (avg)
304.54 tok/s
Prefill (avg)
1,527.8 tok/s
TTFT (avg)
1.17 s
VRAM (avg)
3.41 GB
MTP acceptance rate
—
TTFB
2.44 GB
Power draw
46.3 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-0.8B
Quantization
Q4_K_S
Framework
llama.cpp
Version
v0.41
Context length
8192 tokens
Batch size
256
GPU layers
33
Flash Attention
On

Hardware

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

GPU
Apple M4 Max
Nominal VRAM
128 GB
Measured VRAM (avg)
3.41 GB
OS
macOS 15.0
Driver
Apple M4 Max
CUDA
N/A
Power draw
46.3 W

Sources & evidence

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

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

    v0.41 · macOS 15.0 · CUDA N/A · 8192 ctx

    304.54 tok/s

    Decode

    1,527.8 tok/s

    Prefill

    1.17 s

    TTFT

    3.41 GB

    VRAM

    —

    MTP

    46.3 W

    Power draw

    L2 batch seed 2026-10-10: llama.cpp on hw_id=26 running model_id=51 at Q4_K_S