Arao AI
Back to search results

DeepSeek-R1-Distill-Qwen-14B · Q2_K

Start a discussion
NVIDIA RTX 309024GBvLLM v0.22L2 Cross-framework verified

This page aggregates 1 real-world runs of DeepSeek-R1-Distill-Qwen-14B (Q2_K) on NVIDIA RTX 3090 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

95.91 tok/s

Decode

Decode speed

524.52 tok/s

Prefill

Prefill speed

2.05 s

TTFT

Time to first token

18.27 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. vLLM · Q2_K (current)Decode 95.91 · Prefill 524.52 ·
  2. Ollama · IQ4_XS Decode 73.36 · Prefill 312.95 ·
  3. vLLM · Q5_K_M Decode 76.5 · Prefill 573.8 ·
  4. vLLM · Q4_K_M Decode 93.6 · Prefill 702 ·
  5. llama.cpp · Q5_K_M Decode 70.2 · Prefill 526.5 ·
  6. llama.cpp · Q4_K_M Decode 85.9 · Prefill 644.3 ·

Core figures

Decode (avg)
95.91 tok/s
Prefill (avg)
524.52 tok/s
TTFT (avg)
2.05 s
VRAM (avg)
18.27 GB
MTP acceptance rate
—
TTFB
0.77 GB
Power draw
283.6 W

Configuration

Member-level fields are taken from the most recent run

Model
DeepSeek-R1-Distill-Qwen-14B
Quantization
Q2_K
Framework
vLLM
Version
v0.22
Context length
2048 tokens
Batch size
32
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)
18.27 GB
OS
Ubuntu 24.04
Driver
560.94
CUDA
12.4
Power draw
283.6 W

Sources & evidence

1 measured records in total, each traceable to its original source

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-18

    v0.22 · Ubuntu 24.04 · CUDA 12.4 · 2048 ctx

    95.91 tok/s

    Decode

    524.52 tok/s

    Prefill

    2.05 s

    TTFT

    18.27 GB

    VRAM

    —

    MTP

    283.6 W

    Power draw

    L2 batch seed 2026-10-18: vLLM on hw_id=2 running model_id=71 at Q2_K