← Back to community benchmarks
gpt-oss-120b-MXFP4-Q8
Performance
16k
tokens
766.5
PP tok/s
60.7
TG tok/s
21375
TTFT (ms)
60.4
Peak mem (GB)
Hardware
Chip
M4 Max (40c)
Memory
128 GB
GPU Cores
40
Software
oMLX
v0.3.8
macOS
macOS 26.4.1
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 741.8 | 77.3 | 59.5 GB | view |
| 4k | 839.4 | 73.3 | 59.6 GB | view |
| 8k | 840.7 | 68.1 | 59.9 GB | view |
| 16k | 766.5 | 60.7 | 60.4 GB | current |
| 32k | 672.9 | 49.9 | 61.1 GB | view |
| 64k | 536.3 | 36.3 | 62.8 GB | view |
| 128k | 368.0 | 21.6 | 66.3 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 77.3 | 1.00× |
| 2× | 103.1 | 1.33× |
| 4× | 122.0 | 1.58× |
| 8× | 133.5 | 1.73× |