Arao AI
Back to search results

Qwen3-4B-Thinking-2507 · Q2_K

Start a discussion
NVIDIA Jetson AGX Orin 64GB64GBOllama v0.54L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-4B-Thinking-2507 (Q2_K) on NVIDIA Jetson AGX Orin 64GB with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

77.04 tok/s

Decode

Decode speed

283.31 tok/s

Prefill

Prefill speed

1.72 s

TTFT

Time to first token

6.72 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

2 days ago

Last verified

Performance

  1. vLLM · Q3_K_S Decode 89.8 · Prefill 485.57 ·
  2. vLLM · Q5_K_S Decode 57.15 · Prefill 278.01 ·
  3. Ollama · Q2_K (current)Decode 77.04 · Prefill 283.31 ·
  4. llama.cpp · IQ2_XS Decode 92.94 · Prefill 357.53 ·
  5. Strata · Q5_K_S Decode 83.78 · Prefill 413.16 ·

Core figures

Decode (avg)
77.04 tok/s
Prefill (avg)
283.31 tok/s
TTFT (avg)
1.72 s
VRAM (avg)
6.72 GB
MTP acceptance rate
—
TTFB
0.62 GB
Power draw
22.4 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-4B-Thinking-2507
Quantization
Q2_K
Framework
Ollama
Version
v0.54
Context length
32768 tokens
Batch size
512
GPU layers
20
Flash Attention
On

Hardware

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

GPU
NVIDIA Jetson AGX Orin 64GB
Nominal VRAM
64 GB
Measured VRAM (avg)
6.72 GB
OS
JetPack 6.0
Driver
NVIDIA L4T 36.3
CUDA
12.6
Power draw
22.4 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-16

    v0.54 · JetPack 6.0 · CUDA 12.6 · 32768 ctx

    77.04 tok/s

    Decode

    283.31 tok/s

    Prefill

    1.72 s

    TTFT

    6.72 GB

    VRAM

    —

    MTP

    22.4 W

    Power draw

    L2 batch seed 2026-10-16: Ollama on hw_id=76 running model_id=57 at Q2_K