← Back to community benchmarks

Qwen3.5-35B-A3B

M2 Max (38c) · 64 GB · 8bit · 2026-03-11
Performance
8k
tokens
733.7
PP tok/s
53.2
TG tok/s
11165
TTFT (ms)
36.5
Peak mem (GB)
Hardware
Chip M2 Max (38c)
Memory 64 GB
GPU Cores 38
Software
oMLX v0.2.7
macOS macOS 26.3.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 726.4 61.3 36.2 GB view
4k 746.7 57.7 36.3 GB view
8k 733.7 53.2 36.5 GB current
Batching Results
Batch Size TG tok/s Speedup
61.3 1.00×
101.4 1.65×
148.8 2.43×
165.6 2.70×