Arao AI
Back to search results

gemma-4-E2B-it · Q3_K_S

Start a discussion
Apple M1 MaxvLLM v0.47L2 Cross-framework verified

This page aggregates 1 real-world runs of gemma-4-E2B-it (Q3_K_S) on Apple M1 Max with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

189.55 tok/s

Decode

Decode speed

943.94 tok/s

Prefill

Prefill speed

0.51 s

TTFT

Time to first token

4.83 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 (current)Decode 189.55 · Prefill 943.94 ·
  2. vLLM · Q8_0 Decode 98.82 · Prefill 455.48 ·
  3. Strata · IQ1_S Decode 324.26 · Prefill 1495.83 ·

Core figures

Decode (avg)
189.55 tok/s
Prefill (avg)
943.94 tok/s
TTFT (avg)
0.51 s
VRAM (avg)
4.83 GB
MTP acceptance rate
—
TTFB
1.52 GB
Power draw
58.8 W

Configuration

Member-level fields are taken from the most recent run

Model
gemma-4-E2B-it
Quantization
Q3_K_S
Framework
vLLM
Version
v0.47
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
Apple M1 Max
Measured VRAM (avg)
4.83 GB
OS
macOS 15.0
Driver
Apple M1 Max
CUDA
N/A
Power draw
58.8 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

    v0.47 · macOS 15.0 · CUDA N/A · 32768 ctx

    189.55 tok/s

    Decode

    943.94 tok/s

    Prefill

    0.51 s

    TTFT

    4.83 GB

    VRAM

    —

    MTP

    58.8 W

    Power draw

    L2 batch seed 2026-10-16: vLLM on hw_id=101 running model_id=66 at Q3_K_S