← Back to community benchmarks

gpt-oss-20b-MXFP4-Q8

M3 (10c) · 24 GB · 4bit · 2026-03-11
Performance
8k
tokens
227.1
PP tok/s
22.6
TG tok/s
36066
TTFT (ms)
11.6
Peak mem (GB)
Hardware
Chip M3 (10c)
Memory 24 GB
GPU Cores 10
Software
oMLX v0.2.7
macOS macOS 15.7.3
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 257.2 29.7 11.4 GB view
4k 262.5 26.1 11.5 GB view
8k 227.1 22.6 11.6 GB current
16k 196.2 16.0 11.8 GB view
32k 150.3 2.1 12.2 GB view
Batching Results
Batch Size TG tok/s Speedup
29.7 1.00×
41.3 1.39×
42.5 1.43×
39.4 1.33×