← Back to community benchmarks
gemma-4-E4B-it-The-DECKARD-Expresso-Universe-HERETIC-UNCENSORED-Thinking-mxfp8
Performance
8k
tokens
194.8
PP tok/s
9.7
TG tok/s
42050
TTFT (ms)
8.9
Peak mem (GB)
Hardware
Chip
M1 (7c)
Memory
16 GB
GPU Cores
7
Software
oMLX
v0.3.8
macOS
macOS 26.4.1
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 192.2 | 11.5 | 8.7 GB | view |
| 4k | 201.5 | 11.0 | 8.8 GB | view |
| 8k | 194.8 | 9.7 | 8.9 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 11.5 | 1.00× |
| 2× | 19.4 | 1.69× |
| 4× | 24.4 | 2.12× |