Back to search results
gemma-4-12B-it · IQ4_XS
NVIDIA H100 80GB HBM380GBvLLM v0.24L2 Cross-framework verified
This page aggregates 1 real-world runs of gemma-4-12B-it (IQ4_XS) on NVIDIA H100 80GB HBM3 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.
358.44 tok/s
Decode
Decode speed
1,162.82 tok/s
Prefill
Prefill speed
0.42 s
TTFT
Time to first token
20.75 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
16 days ago
Last verified
Performance
- llama.cpp · Q5_K_M Decode 303.32 · Prefill 1265.92 ·
- llama.cpp · Q5_K_S Decode 283.07 · Prefill 1445.04 ·
- vLLM · IQ3_XS Decode 302.76 · Prefill 1418.76 ·
- vLLM · IQ4_XS (current)Decode 358.44 · Prefill 1162.82 ·
Core figures
- Decode (avg)
- 358.44 tok/s
- Prefill (avg)
- 1,162.82 tok/s
- TTFT (avg)
- 0.42 s
- VRAM (avg)
- 20.75 GB
- MTP acceptance rate
- —
- TTFB
- 2.87 GB
- Power draw
- 613.3 W
Configuration
Member-level fields are taken from the most recent run
- Model
- gemma-4-12B-it
- Quantization
- IQ4_XS
- Framework
- vLLM
- Version
- v0.24
- Context length
- 32768 tokens
- Batch size
- 8
- 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 H100 80GB HBM3
- Nominal VRAM
- 80 GB
- Measured VRAM (avg)
- 20.75 GB
- OS
- Ubuntu 22.04
- Driver
- 560.94
- CUDA
- 12.8
- Power draw
- 613.3 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.24 · Ubuntu 22.04 · CUDA 12.8 · 32768 ctx
358.44 tok/s
Decode
1,162.82 tok/s
Prefill
0.42 s
TTFT
20.75 GB
VRAM
—
MTP
613.3 W
Power draw
L2 batch seed 2026-10-19: vLLM on hw_id=55 running model_id=65 at IQ4_XS