← Back to community benchmarks
gemma-4-26b-a4b
Performance
4k
tokens
1,345
PP tok/s
78.8
TG tok/s
3046
TTFT (ms)
20.5
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.4
macOS
macOS 26.3.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,111 | 83.5 | 20.0 GB | view |
| 4k | 1,345 | 78.8 | 20.5 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 83.5 | 1.00× |
| 2× | 124.7 | 1.49× |
| 4× | 146.3 | 1.75× |