Arao AI
Back to search results

Qwen3-1.7B · IQ2_M

Start a discussion
NVIDIA RTX 507012GBvLLM v0.37L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-1.7B (IQ2_M) on NVIDIA RTX 5070 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

185.93 tok/s

Decode

Decode speed

658.25 tok/s

Prefill

Prefill speed

0.3 s

TTFT

Time to first token

3.96 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. vLLM · IQ2_M (current)Decode 185.93 · Prefill 658.25 ·
  2. Ollama · Q3_K_L Decode 127.9 · Prefill 682.85 ·
  3. Ollama · Q5_K_M Decode 93.03 · Prefill 365.14 ·
  4. Strata · Q3_K_L Decode 187.58 · Prefill 667.52 ·

Core figures

Decode (avg)
185.93 tok/s
Prefill (avg)
658.25 tok/s
TTFT (avg)
0.3 s
VRAM (avg)
3.96 GB
MTP acceptance rate
—
TTFB
1.49 GB
Power draw
150.4 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-1.7B
Quantization
IQ2_M
Framework
vLLM
Version
v0.37
Context length
32768 tokens
Batch size
16
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
NVIDIA RTX 5070
Nominal VRAM
12 GB
Measured VRAM (avg)
3.96 GB
OS
Ubuntu 22.04
Driver
560.94
CUDA
12.8
Power draw
150.4 W

Sources & evidence

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

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

    v0.37 · Ubuntu 22.04 · CUDA 12.8 · 32768 ctx

    185.93 tok/s

    Decode

    658.25 tok/s

    Prefill

    0.3 s

    TTFT

    3.96 GB

    VRAM

    —

    MTP

    150.4 W

    Power draw

    L2 batch seed 2026-10-10: vLLM on hw_id=18 running model_id=55 at IQ2_M