Back to search results
Llama-3.2-3B-Instruct · Q5_K_S
NVIDIA RTX 3060 12GB12GBvLLM v0.45L2 Cross-framework verified
This page aggregates 1 real-world runs of Llama-3.2-3B-Instruct (Q5_K_S) on NVIDIA RTX 3060 12GB with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.
145.02 tok/s
Decode
Decode speed
478.55 tok/s
Prefill
Prefill speed
1.74 s
TTFT
Time to first token
7.77 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
17 days ago
Last verified
Performance
- vLLM · Q4_K_S Decode 182.86 · Prefill 651.97 ·
- vLLM · Q5_K_S (current)Decode 145.02 · Prefill 478.55 ·
- Strata · Q6_K Decode 168.05 · Prefill 844.89 ·
- llama.cpp · IQ3_S Decode 193.44 · Prefill 733.2 ·
- llama.cpp · Q5_K_S Decode 126.84 · Prefill 630.8 ·
- vLLM · Q5_K_M Decode 143.38 · Prefill 521.02 ·
- llama.cpp · Q8_0 Decode 87.8 · Prefill 658.5 ·
- llama.cpp · Q5_K_M Decode 127.6 · Prefill 957 ·
- llama.cpp · Q4_K_M Decode 156 · Prefill 1170 ·
Core figures
- Decode (avg)
- 145.02 tok/s
- Prefill (avg)
- 478.55 tok/s
- TTFT (avg)
- 1.74 s
- VRAM (avg)
- 7.77 GB
- MTP acceptance rate
- —
- TTFB
- 1.16 GB
- Power draw
- 144.4 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Llama-3.2-3B-Instruct
- Quantization
- Q5_K_S
- Framework
- vLLM
- Version
- v0.45
- Context length
- 32768 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
- NVIDIA RTX 3060 12GB
- Nominal VRAM
- 12 GB
- Measured VRAM (avg)
- 7.77 GB
- OS
- Ubuntu 22.04
- Driver
- 565.57
- CUDA
- 12.8
- Power draw
- 144.4 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.45 · Ubuntu 22.04 · CUDA 12.8 · 32768 ctx
145.02 tok/s
Decode
478.55 tok/s
Prefill
1.74 s
TTFT
7.77 GB
VRAM
—
MTP
144.4 W
Power draw
L2 batch seed 2026-10-19: vLLM on hw_id=15 running model_id=70 at Q5_K_S