Arao AI
Back to search results

Qwen3-1.7B · IQ2_M

Start a discussion
Apple M5 Pro64GBllama.cpp v0.50L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-1.7B (IQ2_M) on Apple M5 Pro with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

311.55 tok/s

Decode

Decode speed

1,641.2 tok/s

Prefill

Prefill speed

1.74 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

5 days ago

Last verified

Performance

  1. llama.cpp · IQ2_M (current)Decode 311.55 · Prefill 1641.2 ·
  2. llama.cpp · Q5_K_S Decode 197.87 · Prefill 839.49 ·
  3. vLLM · Q4_K_M Decode 233.37 · Prefill 1190.9 ·
  4. vLLM · Q5_K_S Decode 229.77 · Prefill 936.19 ·

Core figures

Decode (avg)
311.55 tok/s
Prefill (avg)
1,641.2 tok/s
TTFT (avg)
1.74 s
VRAM (avg)
3.96 GB
MTP acceptance rate
—
TTFB
2.49 GB
Power draw
49.9 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-1.7B
Quantization
IQ2_M
Framework
llama.cpp
Version
v0.50
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
Apple M5 Pro
Nominal VRAM
64 GB
Measured VRAM (avg)
3.96 GB
OS
macOS 15.0
Driver
Apple M5 Pro
CUDA
N/A
Power draw
49.9 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

    v0.50 · macOS 15.0 · CUDA N/A · 16384 ctx

    311.55 tok/s

    Decode

    1,641.2 tok/s

    Prefill

    1.74 s

    TTFT

    3.96 GB

    VRAM

    —

    MTP

    49.9 W

    Power draw

    L2 batch seed 2026-10-13: llama.cpp on hw_id=27 running model_id=55 at IQ2_M