← Back to community benchmarks
gemma-4-31B-it-oQ6
Performance
16k
tokens
207.6
PP tok/s
12.5
TG tok/s
78939
TTFT (ms)
28.3
Peak mem (GB)
Hardware
Chip
M1 Ultra (64c)
Memory
128 GB
GPU Cores
64
Software
oMLX
v0.3.8.dev3
macOS
macOS 26.5
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 206.1 | 16.7 | 25.4 GB | view |
| 4k | 215.3 | 15.7 | 27.4 GB | view |
| 8k | 214.7 | 14.5 | 27.7 GB | view |
| 16k | 207.6 | 12.5 | 28.3 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 16.7 | 1.00× |
| 2× | 22.7 | 1.36× |
| 4× | 26.6 | 1.59× |