← Back to community benchmarks
gpt-oss-120b
Performance
4k
tokens
640.8
PP tok/s
59.3
TG tok/s
6392
TTFT (ms)
115.9
Peak mem (GB)
Hardware
Chip
M3 Ultra (80c)
Memory
512 GB
GPU Cores
80
Software
oMLX
v0.2.6
macOS
macOS 26.3.1
Context
4,096
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 528.1 | 63.3 | 115.8 GB | view |
| 4k | 640.8 | 59.3 | 115.9 GB | current |
| 8k | 641.0 | 54.7 | 116.1 GB | view |
| 16k | 622.6 | 47.9 | 116.4 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 63.3 | 1.00× |
| 2× | 87.1 | 1.38× |
| 4× | 113.5 | 1.79× |
| 8× | 132.7 | 2.10× |