← Back to community benchmarks
gemma-4-26B-A4B-it
Performance
4k
tokens
1,321
PP tok/s
98.9
TG tok/s
3101
TTFT (ms)
14.9
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.8
macOS
macOS 15.7.5
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,098 | 99.4 | 14.3 GB | view |
| 4k | 1,321 | 98.9 | 14.9 GB | current |
| 8k | 1,383 | 92.8 | 15.1 GB | view |
| 16k | 1,274 | 83.8 | 15.8 GB | view |
| 32k | 1,068 | 68.8 | 16.5 GB | view |
| 64k | 868.9 | 52.0 | 17.7 GB | view |
| 128k | 546.2 | 29.4 | 21.2 GB | view |
| 195k | 413.9 | 17.5 | 25.0 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 99.4 | 1.00× |
| 2× | 149.5 | 1.50× |
| 4× | 172.5 | 1.74× |
| 8× | 193.2 | 1.94× |