self-hosted
/ai
GPUs
Models
Compare
Tools
Search
⌘K
§01
·
model · /models
Gemma4 31B
llm
active
§02
·
GPUs that run this model
2 total
GPU
VRAM
Series
Best speed
Min VRAM
Works
Benchmarks
RTX 5090
32
GB
50
61.1
tokens/s
—
✓
2
check ↗
RTX 3090
24
GB
30
34.7
tokens/s
24
GB
✓
2
check ↗