← Back to community benchmarks
Qwen3___6-27B
Performance
4k
tokens
413.0
PP tok/s
29.3
TG tok/s
9917
TTFT (ms)
23.5
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
256 GB
GPU Cores
80
Software
oMLX
v0.4.4
macOS
macOS 15.3
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 383.6 | 29.6 | 22.1 GB | view |
| 4k | 413.0 | 29.3 | 23.5 GB | current |
| 8k | 412.2 | 28.7 | 24.2 GB | view |
| 16k | 401.6 | 27.8 | 25.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 29.6 | 1.00× |
| 2× | 48.0 | 1.62× |