← Back to community benchmarks
Qwen3.5-27B-mxfp8
Performance
4k
tokens
181.5
PP tok/s
17.1
TG tok/s
22572
TTFT (ms)
28.7
Peak mem (GB)
Hardware
Chip
M1 Ultra (64c)
Memory
128 GB
GPU Cores
64
Software
oMLX
v0.2.7
macOS
macOS 26.4
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 174.7 | 18.3 | 28.4 GB | view |
| 4k | 181.5 | 17.1 | 28.7 GB | current |
| 8k | 179.6 | 16.7 | 29.2 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 18.3 | 1.00× |
| 2× | 25.9 | 1.42× |
| 4× | 29.0 | 1.58× |