Raon-1-mini-preview · Q5_K_M
This page aggregates 1 real-world runs of Raon-1-mini-preview (Q5_K_M) on NVIDIA H20 with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.
68.61 tok/s
Decode
Decode speed
223.86 tok/s
Prefill
Prefill speed
2.05 s
TTFT
Time to first token
58.35 GB
VRAM
VRAM usage
1
Measured runs
1
Independent sources
Other
Source platforms
13 days ago
Last verified
Performance
- Ollama · IQ4_XS Decode 54.63 · Prefill 291.87 ·
- Ollama · Q3_K_M Decode 58.6 · Prefill 222.5 ·
- Strata · Q5_K_M (current)Decode 68.61 · Prefill 223.86 ·
Core figures
- Decode (avg)
- 68.61 tok/s
- Prefill (avg)
- 223.86 tok/s
- TTFT (avg)
- 2.05 s
- VRAM (avg)
- 58.35 GB
- MTP acceptance rate
- 0.55
- TTFB
- 0.55 GB
- Power draw
- 397.5 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Raon-1-mini-preview
- Quantization
- Q5_K_M
- Framework
- Strata
- Version
- v0.50
- Context length
- 2048 tokens
- Batch size
- 512
- 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 H20
- Nominal VRAM
- 97 GB
- Measured VRAM (avg)
- 58.35 GB
- OS
- Ubuntu 24.04
- Driver
- 565.57
- CUDA
- 12.4
- Power draw
- 397.5 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
v0.50 · Ubuntu 24.04 · CUDA 12.4 · 2048 ctx
68.61 tok/s
Decode
223.86 tok/s
Prefill
2.05 s
TTFT
58.35 GB
VRAM
0.55
MTP
397.5 W
Power draw
L2 batch seed 2026-10-22: Strata on hw_id=56 running model_id=93 at Q5_K_M