Back to search results
Qwen2.5-0.5B-Instruct · IQ3_XS
Qualcomm Snapdragon 7 Gen 4vLLM v0.26L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen2.5-0.5B-Instruct (IQ3_XS) on Qualcomm Snapdragon 7 Gen 4 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.
150.77 tok/s
Decode
Decode speed
624.32 tok/s
Prefill
Prefill speed
1.88 s
TTFT
Time to first token
2.75 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
2 days ago
Last verified
Performance
- vLLM · IQ3_XS (current)Decode 150.77 · Prefill 624.32 ·
- Strata · Q6_K Decode 115.51 · Prefill 492.04 ·
Core figures
- Decode (avg)
- 150.77 tok/s
- Prefill (avg)
- 624.32 tok/s
- TTFT (avg)
- 1.88 s
- VRAM (avg)
- 2.75 GB
- MTP acceptance rate
- —
- TTFB
- 1.21 GB
- Power draw
- 26.7 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen2.5-0.5B-Instruct
- Quantization
- IQ3_XS
- Framework
- vLLM
- Version
- v0.26
- Context length
- 8192 tokens
- Batch size
- 32
- GPU layers
- 999
- Flash Attention
- Off
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- Qualcomm Snapdragon 7 Gen 4
- Measured VRAM (avg)
- 2.75 GB
- OS
- Windows 11 23H2
- Driver
- Qualcomm Adreno 7.0
- CUDA
- N/A
- Power draw
- 26.7 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.26 · Windows 11 23H2 · CUDA N/A · 8192 ctx
150.77 tok/s
Decode
624.32 tok/s
Prefill
1.88 s
TTFT
2.75 GB
VRAM
—
MTP
26.7 W
Power draw
L2 batch seed 2026-10-16: vLLM on hw_id=105 running model_id=59 at IQ3_XS