← Back to community benchmarks
gemma-4-e4b-it-OptiQ
Performance
8k
tokens
2,188
PP tok/s
81.2
TG tok/s
3744
TTFT (ms)
7.0
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
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 | 1,401 | 87.7 | 6.8 GB | view |
| 4k | 1,999 | 84.8 | 6.9 GB | view |
| 8k | 2,188 | 81.2 | 7.0 GB | current |
| 16k | 2,176 | 72.8 | 7.2 GB | view |
| 32k | 1,960 | 61.2 | 7.9 GB | view |
| 64k | 1,590 | 44.9 | 8.8 GB | view |
| 128k | 1,278 | 30.0 | 11.1 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 87.7 | 1.00× |
| 2× | 128.6 | 1.47× |
| 4× | 192.6 | 2.20× |
| 8× | 227.4 | 2.59× |