Back to search results
Qwen3.5-4B · Q5_K_M
NVIDIA RTX 308010GBvLLM v0.32L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen3.5-4B (Q5_K_M) on NVIDIA RTX 3080 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.
82.42 tok/s
Decode
Decode speed
333.84 tok/s
Prefill
Prefill speed
1.83 s
TTFT
Time to first token
9.5 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
5 days ago
Last verified
Performance
- vLLM · Q5_K_M (current)Decode 82.42 · Prefill 333.84 ·
- Ollama · IQ2_M Decode 105.52 · Prefill 563.01 ·
- Ollama · IQ3_S Decode 92.66 · Prefill 391.24 ·
Core figures
- Decode (avg)
- 82.42 tok/s
- Prefill (avg)
- 333.84 tok/s
- TTFT (avg)
- 1.83 s
- VRAM (avg)
- 9.5 GB
- MTP acceptance rate
- —
- TTFB
- 0.66 GB
- Power draw
- 351.5 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3.5-4B
- Quantization
- Q5_K_M
- Framework
- vLLM
- Version
- v0.32
- Context length
- 16384 tokens
- Batch size
- 8
- 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
- NVIDIA RTX 3080
- Nominal VRAM
- 10 GB
- Measured VRAM (avg)
- 9.5 GB
- OS
- Windows 11 23H2
- Driver
- 555.85
- CUDA
- 12.6
- Power draw
- 351.5 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.32 · Windows 11 23H2 · CUDA 12.6 · 16384 ctx
82.42 tok/s
Decode
333.84 tok/s
Prefill
1.83 s
TTFT
9.5 GB
VRAM
—
MTP
351.5 W
Power draw
L2 batch seed 2026-10-13: vLLM on hw_id=44 running model_id=48 at Q5_K_M