← Back to community benchmarks
gemma-4-26B-A4B-it-oQ8
Performance
16k
tokens
3,303
PP tok/s
69.9
TG tok/s
4960
TTFT (ms)
29.2
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.12
macOS
macOS 26.6
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 2,060 | 79.5 | 27.9 GB | view |
| 4k | 2,977 | 76.5 | 28.4 GB | view |
| 8k | 3,255 | 75.2 | 28.8 GB | view |
| 16k | 3,303 | 69.9 | 29.2 GB | current |
| 32k | 2,978 | 62.7 | 30.1 GB | view |
| 64k | 2,379 | 51.9 | 32.1 GB | view |
| 128k | 1,643 | 37.3 | 36.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 79.5 | 1.00× |
| 2× | 120.2 | 1.51× |
| 4× | 165.4 | 2.08× |