Back to search results
DeepSeek-R1-Distill-Qwen-14B · IQ2_XS
NVIDIA RTX 508016GBllama.cpp v0.41L2 Cross-framework verified
This page aggregates 1 real-world runs of DeepSeek-R1-Distill-Qwen-14B (IQ2_XS) on NVIDIA RTX 5080 with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
202.82 tok/s
Decode
Decode speed
808.1 tok/s
Prefill
Prefill speed
1.06 s
TTFT
Time to first token
15.2 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
Today
Last verified
Performance
- llama.cpp · IQ2_XS (current)Decode 202.82 · Prefill 808.1 ·
Core figures
- Decode (avg)
- 202.82 tok/s
- Prefill (avg)
- 808.1 tok/s
- TTFT (avg)
- 1.06 s
- VRAM (avg)
- 15.2 GB
- MTP acceptance rate
- —
- TTFB
- 1.62 GB
- Power draw
- 207.9 W
Configuration
Member-level fields are taken from the most recent run
- Model
- DeepSeek-R1-Distill-Qwen-14B
- Quantization
- IQ2_XS
- Framework
- llama.cpp
- Version
- v0.41
- Context length
- 4096 tokens
- Batch size
- 256
- GPU layers
- 28
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- NVIDIA RTX 5080
- Nominal VRAM
- 16 GB
- Measured VRAM (avg)
- 15.2 GB
- OS
- Windows 11 24H2
- Driver
- 565.57
- CUDA
- 12.8
- Power draw
- 207.9 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.41 · Windows 11 24H2 · CUDA 12.8 · 4096 ctx
202.82 tok/s
Decode
808.1 tok/s
Prefill
1.06 s
TTFT
15.2 GB
VRAM
—
MTP
207.9 W
Power draw
L2 batch seed 2026-10-18: llama.cpp on hw_id=14 running model_id=71 at IQ2_XS