Arao AI
Back to search results

Qwen3.5-4B · IQ2_XS

Start a discussion
AMD Ryzen AI Max+ 395128GBvLLM v0.47L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-4B (IQ2_XS) on AMD Ryzen AI Max+ 395 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

205.56 tok/s

Decode

Decode speed

924.99 tok/s

Prefill

Prefill speed

2.18 s

TTFT

Time to first token

6.13 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 156.15 · Prefill 562.48 ·
  2. vLLM · IQ2_XS (current)Decode 205.56 · Prefill 924.99 ·
  3. vLLM · Q6_K Decode 110.75 · Prefill 500.4 ·

Core figures

Decode (avg)
205.56 tok/s
Prefill (avg)
924.99 tok/s
TTFT (avg)
2.18 s
VRAM (avg)
6.13 GB
MTP acceptance rate
—
TTFB
1.64 GB
Power draw
116.9 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-4B
Quantization
IQ2_XS
Framework
vLLM
Version
v0.47
Context length
2048 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
AMD Ryzen AI Max+ 395
Nominal VRAM
128 GB
Measured VRAM (avg)
6.13 GB
OS
macOS 15.0
Driver
560.94
CUDA
N/A
Power draw
116.9 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 · 2048 ctx

    205.56 tok/s

    Decode

    924.99 tok/s

    Prefill

    2.18 s

    TTFT

    6.13 GB

    VRAM

    —

    MTP

    116.9 W

    Power draw

    L2 batch seed 2026-10-16: vLLM on hw_id=24 running model_id=48 at IQ2_XS