Back to search results
Qwen3-8B · Q3_K_M
NVIDIA RTX 5070 Ti16GBllama.cpp v0.27L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen3-8B (Q3_K_M) on NVIDIA RTX 5070 Ti with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
110.54 tok/s
Decode
Decode speed
601.05 tok/s
Prefill
Prefill speed
0.2 s
TTFT
Time to first token
13.52 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
Today
Last verified
Performance
- llama.cpp · Q3_K_M (current)Decode 110.54 · Prefill 601.05 ·
- Ollama · Q4_K_S Decode 82.54 · Prefill 372.44 ·
Core figures
- Decode (avg)
- 110.54 tok/s
- Prefill (avg)
- 601.05 tok/s
- TTFT (avg)
- 0.2 s
- VRAM (avg)
- 13.52 GB
- MTP acceptance rate
- —
- TTFB
- 0.88 GB
- Power draw
- 226.6 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3-8B
- Quantization
- Q3_K_M
- Framework
- llama.cpp
- Version
- v0.27
- Context length
- 4096 tokens
- Batch size
- 512
- GPU layers
- 20
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- NVIDIA RTX 5070 Ti
- Nominal VRAM
- 16 GB
- Measured VRAM (avg)
- 13.52 GB
- OS
- Ubuntu 22.04
- Driver
- 565.57
- CUDA
- 13.0
- Power draw
- 226.6 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.27 · Ubuntu 22.04 · CUDA 13.0 · 4096 ctx
110.54 tok/s
Decode
601.05 tok/s
Prefill
0.2 s
TTFT
13.52 GB
VRAM
—
MTP
226.6 W
Power draw
L2 batch seed 2026-10-18: llama.cpp on hw_id=17 running model_id=53 at Q3_K_M