← Back to community benchmarks
gemma-4-31b-it
Performance
4k
tokens
147.8
PP tok/s
5.6
TG tok/s
27720
TTFT (ms)
19.2
Peak mem (GB)
Hardware
Chip
M5 (10c)
Memory
32 GB
GPU Cores
10
Software
oMLX
v0.3.8rc1
macOS
macOS 26.3.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 154.9 | 6.1 | 17.4 GB | view |
| 4k | 147.8 | 5.6 | 19.2 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 6.1 | 1.00× |
| 2× | 10.6 | 1.74× |
| 4× | 9.0 | 1.48× |