Arao AI
Back to search results

Qwen3.5-4B · Q4_K_M

Start a discussion
NVIDIA RTX 5070 Ti16GBvLLM v0.29L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-4B (Q4_K_M) on NVIDIA RTX 5070 Ti with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

116.07 tok/s

Decode

Decode speed

492.31 tok/s

Prefill

Prefill speed

0.71 s

TTFT

Time to first token

8.99 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · IQ1_S Decode 182.74 · Prefill 910.54 ·
  2. vLLM · IQ4_XS Decode 131.32 · Prefill 455.29 ·
  3. vLLM · Q4_K_M (current)Decode 116.07 · Prefill 492.31 ·

Core figures

Decode (avg)
116.07 tok/s
Prefill (avg)
492.31 tok/s
TTFT (avg)
0.71 s
VRAM (avg)
8.99 GB
MTP acceptance rate
—
TTFB
0.93 GB
Power draw
235.4 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-4B
Quantization
Q4_K_M
Framework
vLLM
Version
v0.29
Context length
16384 tokens
Batch size
8
GPU layers
999
Flash Attention
Off

Hardware

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

GPU
NVIDIA RTX 5070 Ti
Nominal VRAM
16 GB
Measured VRAM (avg)
8.99 GB
OS
Ubuntu 22.04
Driver
570.86
CUDA
12.8
Power draw
235.4 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-18

    v0.29 · Ubuntu 22.04 · CUDA 12.8 · 16384 ctx

    116.07 tok/s

    Decode

    492.31 tok/s

    Prefill

    0.71 s

    TTFT

    8.99 GB

    VRAM

    —

    MTP

    235.4 W

    Power draw

    L2 batch seed 2026-10-18: vLLM on hw_id=17 running model_id=48 at Q4_K_M