Arao AI
Back to search results

Kimi-K3 · IQ4_XS

Start a discussion
NVIDIA RTX 3060 12GB12GBvLLM v0.21L2 Cross-framework verified

This page aggregates 1 real-world runs of Kimi-K3 (IQ4_XS) on NVIDIA RTX 3060 12GB with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

53.14 tok/s

Decode

Decode speed

228.98 tok/s

Prefill

Prefill speed

2.17 s

TTFT

Time to first token

1.05 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · Q3_K_L Decode 55.93 · Prefill 219.18 ·
  2. llama.cpp · Q3_K_M Decode 58.85 · Prefill 193.51 ·
  3. vLLM · IQ4_XS (current)Decode 53.14 · Prefill 228.98 ·

Core figures

Decode (avg)
53.14 tok/s
Prefill (avg)
228.98 tok/s
TTFT (avg)
2.17 s
VRAM (avg)
1.05 GB
MTP acceptance rate
—
TTFB
0.43 GB
Power draw
168.7 W

Configuration

Member-level fields are taken from the most recent run

Model
Kimi-K3
Quantization
IQ4_XS
Framework
vLLM
Version
v0.21
Context length
4096 tokens
Batch size
8
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)
1.05 GB
OS
Windows 11 24H2
Driver
570.86
CUDA
12.4
Power draw
168.7 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

    v0.21 · Windows 11 24H2 · CUDA 12.4 · 4096 ctx

    53.14 tok/s

    Decode

    228.98 tok/s

    Prefill

    2.17 s

    TTFT

    1.05 GB

    VRAM

    —

    MTP

    168.7 W

    Power draw

    L2 batch seed 2026-10-18: vLLM on hw_id=15 running model_id=73 at IQ4_XS