Back to search results
Qwen2.5-0.5B-Instruct · Q3_K_L
Apple M1 MaxStrata v0.34L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen2.5-0.5B-Instruct (Q3_K_L) on Apple M1 Max with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.
395.33 tok/s
Decode
Decode speed
1,769.04 tok/s
Prefill
Prefill speed
1.45 s
TTFT
Time to first token
2.86 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
2 days ago
Last verified
Performance
- llama.cpp · Q4_K_S Decode 266.38 · Prefill 866.19 ·
- Strata · Q3_K_L (current)Decode 395.33 · Prefill 1769.04 ·
Core figures
- Decode (avg)
- 395.33 tok/s
- Prefill (avg)
- 1,769.04 tok/s
- TTFT (avg)
- 1.45 s
- VRAM (avg)
- 2.86 GB
- MTP acceptance rate
- 0.53
- TTFB
- 3.16 GB
- Power draw
- 55.4 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen2.5-0.5B-Instruct
- Quantization
- Q3_K_L
- Framework
- Strata
- Version
- v0.34
- Context length
- 32768 tokens
- Batch size
- 1024
- GPU layers
- 8
- 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)
- 2.86 GB
- OS
- macOS 15.0
- Driver
- Apple M1 Max
- CUDA
- N/A
- Power draw
- 55.4 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.34 · macOS 15.0 · CUDA N/A · 32768 ctx
395.33 tok/s
Decode
1,769.04 tok/s
Prefill
1.45 s
TTFT
2.86 GB
VRAM
0.53
MTP
55.4 W
Power draw
L2 batch seed 2026-10-16: Strata on hw_id=101 running model_id=59 at Q3_K_L