← Back to community benchmarks
gemma-4-26B-A4B-it-oQ8
Performance
32k
tokens
1,953
PP tok/s
60.3
TG tok/s
16781
TTFT (ms)
27.5
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.3.10
macOS
macOS 26.5
Context
32,768
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 1,423 | 80.3 | 25.8 GB | view |
| 4k | 2,092 | 78.3 | 26.1 GB | view |
| 8k | 2,178 | 75.1 | 26.3 GB | view |
| 16k | 2,148 | 70.0 | 26.7 GB | view |
| 32k | 1,953 | 60.3 | 27.5 GB | current |
| 64k | 1,584 | 45.8 | 29.2 GB | view |
| 128k | 1,105 | 30.6 | 32.7 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 80.3 | 1.00× |
| 2× | 125.1 | 1.56× |
| 4× | 180.3 | 2.25× |
| 8× | 252.9 | 3.15× |