← Back to community benchmarks

Qwen3.5-35B-A3B

M4 Max (40c) · 64 GB · 4bit · 2026-03-13
Performance
1k
tokens
1,732
PP tok/s
114.4
TG tok/s
591
TTFT (ms)
20.1
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 64 GB
GPU Cores 40
Software
oMLX v0.2.10
macOS macOS 26.3.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,732 114.4 20.1 GB current
4k 1,755 108.5 20.2 GB view
8k 1,677 93.7 20.4 GB view
Batching Results
Batch Size TG tok/s Speedup
114.4 1.00×
174.5 1.53×
272.5 2.38×
346.5 3.03×