Arao AI
Back to search results

Qwen3-4B-Thinking-2507 · Q3_K_M

Start a discussion
AMD Radeon RX 9070 XT16GBOllama v0.36L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-4B-Thinking-2507 (Q3_K_M) on AMD Radeon RX 9070 XT with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

127.72 tok/s

Decode

Decode speed

433.24 tok/s

Prefill

Prefill speed

1.09 s

TTFT

Time to first token

7.81 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 · IQ2_XS Decode 186.6 · Prefill 815.85 ·
  2. vLLM · Q5_K_S Decode 103.28 · Prefill 532.39 ·
  3. Ollama · Q3_K_M (current)Decode 127.72 · Prefill 433.24 ·
  4. Ollama · Q3_K_S Decode 126.74 · Prefill 598.67 ·

Core figures

Decode (avg)
127.72 tok/s
Prefill (avg)
433.24 tok/s
TTFT (avg)
1.09 s
VRAM (avg)
7.81 GB
MTP acceptance rate
—
TTFB
1.02 GB
Power draw
318.2 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-4B-Thinking-2507
Quantization
Q3_K_M
Framework
Ollama
Version
v0.36
Context length
16384 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
AMD Radeon RX 9070 XT
Nominal VRAM
16 GB
Measured VRAM (avg)
7.81 GB
OS
Ubuntu 22.04
Driver
amdgpu-pro 24.20
CUDA
N/A
Power draw
318.2 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

    v0.36 · Ubuntu 22.04 · CUDA N/A · 16384 ctx

    127.72 tok/s

    Decode

    433.24 tok/s

    Prefill

    1.09 s

    TTFT

    7.81 GB

    VRAM

    —

    MTP

    318.2 W

    Power draw

    L2 batch seed 2026-10-13: Ollama on hw_id=16 running model_id=57 at Q3_K_M