← Back to community benchmarks
Qwen3.5-27B
Performance
8k
tokens
275.6
PP tok/s
8.8
TG tok/s
29721
TTFT (ms)
54.9
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.19
macOS
macOS 26.3
Context
8,192
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 268.2 | 9.3 | 52.5 GB | view |
| 4k | 278.2 | 9.1 | 54.2 GB | view |
| 8k | 275.6 | 8.8 | 54.9 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 9.3 | 1.00× |
| 2× | 7.9 | 0.85× |
| 4× | 15.5 | 1.67× |
| 8× | 30.3 | 3.26× |