← Back to community benchmarks
gemma-4-31b
Performance
16k
tokens
240.5
PP tok/s
10.1
TG tok/s
68119
TTFT (ms)
63.0
Peak mem (GB)
Hardware
Chip
M3 Ultra (60c)
Memory
256 GB
GPU Cores
60
Software
oMLX
v0.3.8
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 223.4 | 10.7 | 58.9 GB | view |
| 4k | 247.1 | 10.6 | 61.2 GB | view |
| 8k | 248.2 | 10.5 | 61.7 GB | view |
| 16k | 240.5 | 10.1 | 63.0 GB | current |
| 32k | 224.1 | 9.4 | 65.8 GB | view |
| 64k | 194.2 | 8.0 | 71.8 GB | view |
| 128k | 149.7 | 6.6 | 83.9 GB | view |
| 195k | 120.3 | 5.3 | 98.9 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 10.7 | 1.00× |
| 2× | 9.3 | 0.87× |
| 4× | 17.5 | 1.64× |
| 8× | 34.2 | 3.20× |