← Back to community benchmarks

Ling-2.6-flash4bit-DWQ

M2 Max (38c) · 96 GB · 4bit · 2026-04-30
Performance
8k
tokens
251.6
PP tok/s
37.2
TG tok/s
32563
TTFT (ms)
56.6
Peak mem (GB)
Hardware
Chip M2 Max (38c)
Memory 96 GB
GPU Cores 38
Software
oMLX v0.3.8
macOS macOS 15.7.4
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 291.5 36.9 55.3 GB view
4k 249.8 39.2 56.0 GB view
8k 251.6 37.2 56.6 GB current
16k 246.9 36.3 57.8 GB view
32k 239.7 34.8 60.1 GB view
64k 220.8 30.8 64.8 GB view
Batching Results
Batch Size TG tok/s Speedup
36.9 1.00×
46.1 1.25×
51.4 1.39×