← Back to community benchmarks
gemma-4-26B-A4B-it-oQ8
Performance
8k
tokens
3,331
PP tok/s
74.5
TG tok/s
2459
TTFT (ms)
26.4
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.6
macOS
macOS 26.5
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,041 | 85.7 | 25.7 GB | view |
| 4k | 3,056 | 80.4 | 26.2 GB | view |
| 8k | 3,331 | 74.5 | 26.4 GB | current |
| 16k | 3,319 | 63.0 | 26.7 GB | view |
| 32k | 3,016 | 43.3 | 27.5 GB | view |
| 64k | 2,450 | 18.4 | 29.2 GB | view |
| 128k | 1,705 | 8.1 | 32.7 GB | view |
| 195k | 1,268 | 5.5 | 36.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 85.7 | 1.00× |
| 2× | 119.6 | 1.40× |
| 4× | 174.9 | 2.04× |
| 8× | 347.6 | 4.06× |