Arao AI
Back to search results

Kimi-K3 · FP16

Start a discussion
NVIDIA RTX 408016GBStrata v0.40L2 Cross-framework verified

This page aggregates 1 real-world runs of Kimi-K3 (FP16) on NVIDIA RTX 4080 with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.

79.66 tok/s

Decode

Decode speed

266.23 tok/s

Prefill

Prefill speed

2.28 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. Ollama · IQ3_XS Decode 139.12 · Prefill 753.5 ·
  2. Ollama · Q5_K_M Decode 91.51 · Prefill 334.92 ·
  3. Strata · FP16 (current)Decode 79.66 · Prefill 266.23 ·
  4. Strata · Q4_K_M Decode 165.81 · Prefill 742.77 ·

Core figures

Decode (avg)
79.66 tok/s
Prefill (avg)
266.23 tok/s
TTFT (avg)
2.28 s
VRAM (avg)
1.05 GB
MTP acceptance rate
0.53
TTFB
0.64 GB
Power draw
279 W

Configuration

Member-level fields are taken from the most recent run

Model
Kimi-K3
Quantization
FP16
Framework
Strata
Version
v0.40
Context length
32768 tokens
Batch size
256
GPU layers
15
Flash Attention
Off

Hardware

Nominal and measured figures are shown side by side; whether it runs is the reader's call

GPU
NVIDIA RTX 4080
Nominal VRAM
16 GB
Measured VRAM (avg)
1.05 GB
OS
Windows 11 24H2
Driver
565.57
CUDA
13.0
Power draw
279 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

    v0.40 · Windows 11 24H2 · CUDA 13.0 · 32768 ctx

    79.66 tok/s

    Decode

    266.23 tok/s

    Prefill

    2.28 s

    TTFT

    1.05 GB

    VRAM

    0.53

    MTP

    279 W

    Power draw

    L2 batch seed 2026-10-18: Strata on hw_id=41 running model_id=73 at FP16