← Back to community benchmarks
gemma-4-31b-it
Performance
8k
tokens
249.9
PP tok/s
17.3
TG tok/s
32784
TTFT (ms)
35.0
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.3.12
macOS
macOS 26.5
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 243.9 | 17.1 | 31.8 GB | view |
| 4k | 250.9 | 17.9 | 34.5 GB | view |
| 8k | 249.9 | 17.3 | 35.0 GB | current |
| 16k | 240.2 | 16.2 | 36.1 GB | view |
| 32k | 221.8 | 13.5 | 39.2 GB | view |
| 64k | 193.3 | 11.4 | 45.8 GB | view |
| 128k | 149.6 | 8.8 | 57.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 17.1 | 1.00× |
| 2× | 29.7 | 1.74× |
| 4× | 37.2 | 2.18× |