← Back to community benchmarks
gpt-oss-120b-MXFP4-Q8
Performance
64k
tokens
514.0
PP tok/s
16.3
TG tok/s
127512
TTFT (ms)
61.6
Peak mem (GB)
Hardware
Chip
M5 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.2.10
macOS
macOS 26.3.1
Context
65,536
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 497.6 | 85.1 | 59.2 GB | view |
| 4k | 619.3 | 77.3 | 59.3 GB | view |
| 8k | 639.2 | 63.5 | 59.5 GB | view |
| 16k | 622.3 | 38.4 | 59.8 GB | view |
| 64k | 514.0 | 16.3 | 61.6 GB | current |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 85.1 | 1.00× |
| 2× | 118.2 | 1.39× |
| 4× | 145.2 | 1.71× |
| 8× | 157.2 | 1.85× |