Back to search results
Qwen3-4B-Thinking-2507 · Q4_K_M
NVIDIA RTX 4070 Ti12GBvLLM v0.28L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen3-4B-Thinking-2507 (Q4_K_M) on NVIDIA RTX 4070 Ti with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.
99.74 tok/s
Decode
Decode speed
485.1 tok/s
Prefill
Prefill speed
0.2 s
TTFT
Time to first token
8.99 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
5 days ago
Last verified
Performance
- vLLM · Q4_K_M (current)Decode 99.74 · Prefill 485.1 ·
- vLLM · Q8_0 Decode 64.87 · Prefill 300.97 ·
Core figures
- Decode (avg)
- 99.74 tok/s
- Prefill (avg)
- 485.1 tok/s
- TTFT (avg)
- 0.2 s
- VRAM (avg)
- 8.99 GB
- MTP acceptance rate
- —
- TTFB
- 0.8 GB
- Power draw
- 252.1 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3-4B-Thinking-2507
- Quantization
- Q4_K_M
- Framework
- vLLM
- Version
- v0.28
- Context length
- 8192 tokens
- Batch size
- 16
- 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
- NVIDIA RTX 4070 Ti
- Nominal VRAM
- 12 GB
- Measured VRAM (avg)
- 8.99 GB
- OS
- Ubuntu 24.04
- Driver
- 555.85
- CUDA
- 12.6
- Power draw
- 252.1 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.28 · Ubuntu 24.04 · CUDA 12.6 · 8192 ctx
99.74 tok/s
Decode
485.1 tok/s
Prefill
0.2 s
TTFT
8.99 GB
VRAM
—
MTP
252.1 W
Power draw
L2 batch seed 2026-10-13: vLLM on hw_id=42 running model_id=57 at Q4_K_M