Arao AI
Back to search results

Llama-3.1-8B-Instruct · Q4_K_M

Start a discussion
NVIDIA RTX 409024GBvLLM v0.29L2 Cross-framework verified

This page aggregates 1 real-world runs of Llama-3.1-8B-Instruct (Q4_K_M) on NVIDIA RTX 4090 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

142.74 tok/s

Decode

Decode speed

510.45 tok/s

Prefill

Prefill speed

2.44 s

TTFT

Time to first token

15.88 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · Q5_K_M Decode 128.55 · Prefill 827.45 ·
  2. vLLM · Q4_K_M (current)Decode 142.74 · Prefill 510.45 ·
  3. llama.cpp · Q8_0 Decode 91.4 · Prefill 685.5 ·
  4. llama.cpp · Q4_K_M Decode 160.5 · Prefill 1203.8 ·

Core figures

Decode (avg)
142.74 tok/s
Prefill (avg)
510.45 tok/s
TTFT (avg)
2.44 s
VRAM (avg)
15.88 GB
MTP acceptance rate
—
TTFB
1.14 GB
Power draw
393.3 W

Configuration

Member-level fields are taken from the most recent run

Model
Llama-3.1-8B-Instruct
Quantization
Q4_K_M
Framework
vLLM
Version
v0.29
Context length
8192 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 4090
Nominal VRAM
24 GB
Measured VRAM (avg)
15.88 GB
OS
Windows 11 24H2
Driver
570.86
CUDA
12.8
Power draw
393.3 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

    v0.29 · Windows 11 24H2 · CUDA 12.8 · 8192 ctx

    142.74 tok/s

    Decode

    510.45 tok/s

    Prefill

    2.44 s

    TTFT

    15.88 GB

    VRAM

    —

    MTP

    393.3 W

    Power draw

    L2 batch seed 2026-10-18: vLLM on hw_id=1 running model_id=68 at Q4_K_M