Back to search results
DeepSeek-R1-Distill-Qwen-14B · Q4_K_S
Apple M2 Ultrallama.cpp v0.37L2 Cross-framework verified
This page aggregates 1 real-world runs of DeepSeek-R1-Distill-Qwen-14B (Q4_K_S) on Apple M2 Ultra with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
92.27 tok/s
Decode
Decode speed
397.78 tok/s
Prefill
Prefill speed
0.42 s
TTFT
Time to first token
25.03 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
16 days ago
Last verified
Performance
- llama.cpp · Q4_K_S (current)Decode 92.27 · Prefill 397.78 ·
- Ollama · Q3_K_S Decode 106.35 · Prefill 407.93 ·
- Ollama · Q5_K_M Decode 66.35 · Prefill 310.63 ·
Core figures
- Decode (avg)
- 92.27 tok/s
- Prefill (avg)
- 397.78 tok/s
- TTFT (avg)
- 0.42 s
- VRAM (avg)
- 25.03 GB
- MTP acceptance rate
- —
- TTFB
- 0.74 GB
- Power draw
- 130.9 W
Configuration
Member-level fields are taken from the most recent run
- Model
- DeepSeek-R1-Distill-Qwen-14B
- Quantization
- Q4_K_S
- Framework
- llama.cpp
- Version
- v0.37
- Context length
- 8192 tokens
- Batch size
- 256
- GPU layers
- 33
- 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)
- 25.03 GB
- OS
- macOS 26.1
- Power draw
- 130.9 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.37 · macOS 26.1 · 8192 ctx
92.27 tok/s
Decode
397.78 tok/s
Prefill
0.42 s
TTFT
25.03 GB
VRAM
—
MTP
130.9 W
Power draw
L2 batch seed 2026-10-19: llama.cpp on hw_id=103 running model_id=71 at Q4_K_S