Arao AI
Back to search results

DeepSeek-R1-Distill-Qwen-14B · Q3_K_S

Start a discussion
Apple M2 UltraOllama v0.44L2 Cross-framework verified

This page aggregates 1 real-world runs of DeepSeek-R1-Distill-Qwen-14B (Q3_K_S) on Apple M2 Ultra with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

106.35 tok/s

Decode

Decode speed

407.93 tok/s

Prefill

Prefill speed

0.88 s

TTFT

Time to first token

21.21 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

17 days ago

Last verified

Performance

  1. llama.cpp · Q4_K_S Decode 92.27 · Prefill 397.78 ·
  2. Ollama · Q3_K_S (current)Decode 106.35 · Prefill 407.93 ·
  3. Ollama · Q5_K_M Decode 66.35 · Prefill 310.63 ·

Core figures

Decode (avg)
106.35 tok/s
Prefill (avg)
407.93 tok/s
TTFT (avg)
0.88 s
VRAM (avg)
21.21 GB
MTP acceptance rate
—
TTFB
0.85 GB
Power draw
149.9 W

Configuration

Member-level fields are taken from the most recent run

Model
DeepSeek-R1-Distill-Qwen-14B
Quantization
Q3_K_S
Framework
Ollama
Version
v0.44
Context length
8192 tokens
Batch size
256
GPU layers
28
Flash Attention
On

Hardware

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

GPU
Apple M2 Ultra
Measured VRAM (avg)
21.21 GB
OS
macOS 26.1
Power draw
149.9 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-19

    v0.44 · macOS 26.1 · 8192 ctx

    106.35 tok/s

    Decode

    407.93 tok/s

    Prefill

    0.88 s

    TTFT

    21.21 GB

    VRAM

    —

    MTP

    149.9 W

    Power draw

    L2 batch seed 2026-10-19: Ollama on hw_id=103 running model_id=71 at Q3_K_S