Back to search results
Qwen3.5-4B · Q5_K_S
NVIDIA RTX 309024GBvLLM v0.46L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen3.5-4B (Q5_K_S) on NVIDIA RTX 3090 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.
129.61 tok/s
Decode
Decode speed
696.02 tok/s
Prefill
Prefill speed
0.96 s
TTFT
Time to first token
9.66 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
Today
Last verified
Performance
- llama.cpp · Q5_K_S Decode 139.06 · Prefill 548.16 ·
- llama.cpp · Q6_K Decode 108.41 · Prefill 420.36 ·
- vLLM · Q5_K_S (current)Decode 129.61 · Prefill 696.02 ·
Core figures
- Decode (avg)
- 129.61 tok/s
- Prefill (avg)
- 696.02 tok/s
- TTFT (avg)
- 0.96 s
- VRAM (avg)
- 9.66 GB
- MTP acceptance rate
- —
- TTFB
- 1.04 GB
- Power draw
- 314.8 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3.5-4B
- Quantization
- Q5_K_S
- Framework
- vLLM
- Version
- v0.46
- 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 3090
- Nominal VRAM
- 24 GB
- Measured VRAM (avg)
- 9.66 GB
- OS
- Ubuntu 24.04
- Driver
- 560.94
- CUDA
- 12.8
- Power draw
- 314.8 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.46 · Ubuntu 24.04 · CUDA 12.8 · 16384 ctx
129.61 tok/s
Decode
696.02 tok/s
Prefill
0.96 s
TTFT
9.66 GB
VRAM
—
MTP
314.8 W
Power draw
L2 batch seed 2026-10-10: vLLM on hw_id=2 running model_id=48 at Q5_K_S