Arao AI
Back to search results

Qwen3.5-0.8B · IQ4_XS

Start a discussion
Apple M4 Pro64GBllama.cpp v0.52L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-0.8B (IQ4_XS) on Apple M4 Pro with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

286.93 tok/s

Decode

Decode speed

1,431.45 tok/s

Prefill

Prefill speed

2.26 s

TTFT

Time to first token

3.34 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

5 days ago

Last verified

Performance

  1. llama.cpp · IQ4_XS (current)Decode 286.93 · Prefill 1431.45 ·
  2. llama.cpp · Q5_K_M Decode 205.09 · Prefill 1101.85 ·

Core figures

Decode (avg)
286.93 tok/s
Prefill (avg)
1,431.45 tok/s
TTFT (avg)
2.26 s
VRAM (avg)
3.34 GB
MTP acceptance rate
—
TTFB
2.3 GB
Power draw
114.7 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-0.8B
Quantization
IQ4_XS
Framework
llama.cpp
Version
v0.52
Context length
32768 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 Pro
Nominal VRAM
64 GB
Measured VRAM (avg)
3.34 GB
OS
macOS 15.0
Driver
Apple M4 Pro
CUDA
N/A
Power draw
114.7 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-13

    v0.52 · macOS 15.0 · CUDA N/A · 32768 ctx

    286.93 tok/s

    Decode

    1,431.45 tok/s

    Prefill

    2.26 s

    TTFT

    3.34 GB

    VRAM

    —

    MTP

    114.7 W

    Power draw

    L2 batch seed 2026-10-13: llama.cpp on hw_id=25 running model_id=51 at IQ4_XS