Arao AI
Back to search results

Raon-1-mini-preview · IQ4_XS

Start a discussion
NVIDIA H2097GBOllama v0.21L2 Cross-framework verified

This page aggregates 1 real-world runs of Raon-1-mini-preview (IQ4_XS) on NVIDIA H20 with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

54.63 tok/s

Decode

Decode speed

291.87 tok/s

Prefill

Prefill speed

2.29 s

TTFT

Time to first token

46.85 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

13 days ago

Last verified

Performance

  1. Ollama · IQ4_XS (current)Decode 54.63 · Prefill 291.87 ·
  2. Ollama · Q3_K_M Decode 58.6 · Prefill 222.5 ·
  3. Strata · Q5_K_M Decode 68.61 · Prefill 223.86 ·

Core figures

Decode (avg)
54.63 tok/s
Prefill (avg)
291.87 tok/s
TTFT (avg)
2.29 s
VRAM (avg)
46.85 GB
MTP acceptance rate
—
TTFB
0.44 GB
Power draw
337.7 W

Configuration

Member-level fields are taken from the most recent run

Model
Raon-1-mini-preview
Quantization
IQ4_XS
Framework
Ollama
Version
v0.21
Context length
2048 tokens
Batch size
512
GPU layers
99
Flash Attention
On

Hardware

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

GPU
NVIDIA H20
Nominal VRAM
97 GB
Measured VRAM (avg)
46.85 GB
OS
Windows 11 23H2
Driver
570.86
CUDA
12.6
Power draw
337.7 W

Get started

Original model resources

Base model resources may not match this quantization. Use the verified deployment weights above when available.

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-22

    v0.21 · Windows 11 23H2 · CUDA 12.6 · 2048 ctx

    54.63 tok/s

    Decode

    291.87 tok/s

    Prefill

    2.29 s

    TTFT

    46.85 GB

    VRAM

    —

    MTP

    337.7 W

    Power draw

    L2 batch seed 2026-10-22: Ollama on hw_id=56 running model_id=93 at IQ4_XS