← Back to community benchmarks
gemma-4-26B-A4B-it-oQ8
Performance
16k
tokens
1,335
PP tok/s
58.0
TG tok/s
12273
TTFT (ms)
26.7
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.6
macOS
macOS 15.7.4
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,170 | 75.6 | 25.7 GB | view |
| 4k | 1,363 | 71.6 | 26.2 GB | view |
| 8k | 1,383 | 66.6 | 26.3 GB | view |
| 16k | 1,335 | 58.0 | 26.7 GB | current |
| 32k | 1,207 | 46.3 | 27.5 GB | view |
| 64k | 1,009 | 24.3 | 29.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 75.6 | 1.00× |
| 2× | 111.9 | 1.48× |
| 4× | 154.7 | 2.05× |
| 8× | 198.1 | 2.62× |