Back to search results
Qwen3-4B-Thinking-2507 · Q3_K_L
Apple M5 Pro64GBllama.cpp v0.41L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen3-4B-Thinking-2507 (Q3_K_L) on Apple M5 Pro with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
181 tok/s
Decode
Decode speed
762.88 tok/s
Prefill
Prefill speed
1.31 s
TTFT
Time to first token
8.15 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
5 days ago
Last verified
Performance
- llama.cpp · Q3_K_L (current)Decode 181 · Prefill 762.88 ·
Core figures
- Decode (avg)
- 181 tok/s
- Prefill (avg)
- 762.88 tok/s
- TTFT (avg)
- 1.31 s
- VRAM (avg)
- 8.15 GB
- MTP acceptance rate
- —
- TTFB
- 1.45 GB
- Power draw
- 132.2 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3-4B-Thinking-2507
- Quantization
- Q3_K_L
- Framework
- llama.cpp
- Version
- v0.41
- Context length
- 8192 tokens
- Batch size
- 512
- GPU layers
- 20
- Flash Attention
- Off
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)
- 8.15 GB
- OS
- macOS 15.0
- Driver
- Apple M5 Pro
- CUDA
- N/A
- Power draw
- 132.2 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.41 · macOS 15.0 · CUDA N/A · 8192 ctx
181 tok/s
Decode
762.88 tok/s
Prefill
1.31 s
TTFT
8.15 GB
VRAM
—
MTP
132.2 W
Power draw
L2 batch seed 2026-10-13: llama.cpp on hw_id=27 running model_id=57 at Q3_K_L