Arao AI
Back to search results

Qwen2.5-Omni-3B · IQ4_XS

Start a discussion
NVIDIA DGX Spark128GBOllama v0.52L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen2.5-Omni-3B (IQ4_XS) on NVIDIA DGX Spark with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

218.18 tok/s

Decode

Decode speed

903.32 tok/s

Prefill

Prefill speed

1.07 s

TTFT

Time to first token

6.76 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

5 days ago

Last verified

Performance

  1. vLLM · Q3_K_M Decode 258.63 · Prefill 1322.45 ·
  2. vLLM · Q5_K_M Decode 183.75 · Prefill 675.77 ·
  3. Ollama · IQ2_M Decode 292.67 · Prefill 1302.54 ·
  4. Ollama · IQ4_XS (current)Decode 218.18 · Prefill 903.32 ·

Core figures

Decode (avg)
218.18 tok/s
Prefill (avg)
903.32 tok/s
TTFT (avg)
1.07 s
VRAM (avg)
6.76 GB
MTP acceptance rate
—
TTFB
1.75 GB
Power draw
219.6 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen2.5-Omni-3B
Quantization
IQ4_XS
Framework
Ollama
Version
v0.52
Context length
4096 tokens
Batch size
256
GPU layers
99
Flash Attention
On

Hardware

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

GPU
NVIDIA DGX Spark
Nominal VRAM
128 GB
Measured VRAM (avg)
6.76 GB
OS
Ubuntu 24.04
Driver
565.57
CUDA
12.8
Power draw
219.6 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 · Ubuntu 24.04 · CUDA 12.8 · 4096 ctx

    218.18 tok/s

    Decode

    903.32 tok/s

    Prefill

    1.07 s

    TTFT

    6.76 GB

    VRAM

    —

    MTP

    219.6 W

    Power draw

    L2 batch seed 2026-10-13: Ollama on hw_id=54 running model_id=60 at IQ4_XS