← Back to community benchmarks
gpt-oss-20b-MXFP4-Q8
Performance
1k
tokens
773.6
PP tok/s
37.8
TG tok/s
1324
TTFT (ms)
11.7
Peak mem (GB)
Hardware
Chip
M5 (10c)
Memory
24 GB
GPU Cores
10
Software
oMLX
v0.3.0
macOS
macOS 26.4.1
Context
1,024
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 773.6 | 37.8 | 11.7 GB | current |
| 4k | 1,087 | 35.9 | 11.8 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 37.8 | 1.00× |
| 2× | 52.8 | 1.40× |
| 4× | 61.5 | 1.63× |
| 8× | 57.9 | 1.53× |