Back to search results
Qwen3.5-0.8B · IQ3_XS
Apple M1 ProvLLM v0.55L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen3.5-0.8B (IQ3_XS) on Apple M1 Pro with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.
221.94 tok/s
Decode
Decode speed
834.89 tok/s
Prefill
Prefill speed
1.34 s
TTFT
Time to first token
3.14 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
2 days ago
Last verified
Performance
- vLLM · FP16 Decode 88.25 · Prefill 446.82 ·
- vLLM · IQ3_XS (current)Decode 221.94 · Prefill 834.89 ·
- Strata · IQ4_XS Decode 263.91 · Prefill 1056.07 ·
- Strata · Q6_K Decode 165.63 · Prefill 531.23 ·
Core figures
- Decode (avg)
- 221.94 tok/s
- Prefill (avg)
- 834.89 tok/s
- TTFT (avg)
- 1.34 s
- VRAM (avg)
- 3.14 GB
- MTP acceptance rate
- —
- TTFB
- 1.78 GB
- Power draw
- 44.9 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3.5-0.8B
- Quantization
- IQ3_XS
- Framework
- vLLM
- Version
- v0.55
- Context length
- 16384 tokens
- Batch size
- 16
- GPU layers
- 999
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- Apple M1 Pro
- Measured VRAM (avg)
- 3.14 GB
- OS
- macOS 15.0
- Driver
- Apple M1 Pro
- CUDA
- N/A
- Power draw
- 44.9 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.55 · macOS 15.0 · CUDA N/A · 16384 ctx
221.94 tok/s
Decode
834.89 tok/s
Prefill
1.34 s
TTFT
3.14 GB
VRAM
—
MTP
44.9 W
Power draw
L2 batch seed 2026-10-16: vLLM on hw_id=100 running model_id=51 at IQ3_XS