Back to search results
Qwen3-4B-Thinking-2507 · IQ3_XS
Apple M4 Pro64GBStrata v0.40L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen3-4B-Thinking-2507 (IQ3_XS) on Apple M4 Pro with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.
218.39 tok/s
Decode
Decode speed
1,145.06 tok/s
Prefill
Prefill speed
1.2 s
TTFT
Time to first token
7.31 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 · IQ4_XS Decode 138.37 · Prefill 661.86 ·
- llama.cpp · Q4_K_S Decode 123.54 · Prefill 441.3 ·
- Strata · IQ3_XS (current)Decode 218.39 · Prefill 1145.06 ·
- Strata · Q3_K_L Decode 206.28 · Prefill 783.16 ·
Core figures
- Decode (avg)
- 218.39 tok/s
- Prefill (avg)
- 1,145.06 tok/s
- TTFT (avg)
- 1.2 s
- VRAM (avg)
- 7.31 GB
- MTP acceptance rate
- 0.59
- TTFB
- 1.75 GB
- Power draw
- 110.8 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3-4B-Thinking-2507
- Quantization
- IQ3_XS
- Framework
- Strata
- Version
- v0.40
- Context length
- 16384 tokens
- Batch size
- 256
- GPU layers
- 15
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- Apple M4 Pro
- Nominal VRAM
- 64 GB
- Measured VRAM (avg)
- 7.31 GB
- OS
- macOS 15.0
- Driver
- Apple M4 Pro
- CUDA
- N/A
- Power draw
- 110.8 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.40 · macOS 15.0 · CUDA N/A · 16384 ctx
218.39 tok/s
Decode
1,145.06 tok/s
Prefill
1.2 s
TTFT
7.31 GB
VRAM
0.59
MTP
110.8 W
Power draw
L2 batch seed 2026-10-13: Strata on hw_id=25 running model_id=57 at IQ3_XS