Arao AI
Back to search results

Qwen3-Embedding-0.6B · IQ2_M

Start a discussion
Apple M5 Max128GBvLLM v0.32L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-Embedding-0.6B (IQ2_M) on Apple M5 Max with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

302.89 tok/s

Decode

Decode speed

1,094.91 tok/s

Prefill

Prefill speed

2 s

TTFT

Time to first token

2.75 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

2 days ago

Last verified

Performance

  1. llama.cpp · Q6_K Decode 297.37 · Prefill 1423.38 ·
  2. vLLM · FP16 Decode 257.66 · Prefill 966.33 ·
  3. vLLM · IQ2_M (current)Decode 302.89 · Prefill 1094.91 ·
  4. Ollama · IQ3_XS Decode 290.71 · Prefill 1516.13 ·

Core figures

Decode (avg)
302.89 tok/s
Prefill (avg)
1,094.91 tok/s
TTFT (avg)
2 s
VRAM (avg)
2.75 GB
MTP acceptance rate
—
TTFB
2.42 GB
Power draw
223.9 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-Embedding-0.6B
Quantization
IQ2_M
Framework
vLLM
Version
v0.32
Context length
4096 tokens
Batch size
8
GPU layers
999
Flash Attention
Off

Hardware

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

GPU
Apple M5 Max
Nominal VRAM
128 GB
Measured VRAM (avg)
2.75 GB
OS
macOS 15.0
Driver
555.85
CUDA
12.8
Power draw
223.9 W

Sources & evidence

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

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

    v0.32 · macOS 15.0 · CUDA 12.8 · 4096 ctx

    302.89 tok/s

    Decode

    1,094.91 tok/s

    Prefill

    2 s

    TTFT

    2.75 GB

    VRAM

    —

    MTP

    223.9 W

    Power draw

    L2 batch seed 2026-10-16: vLLM on hw_id=28 running model_id=62 at IQ2_M