← Back to community benchmarks

Qwen3.5-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled-text-oQ3.5

M4 Max (40c) · 128 GB · 3bit · 2026-04-16
Performance
8k
tokens
1,681
PP tok/s
99.1
TG tok/s
4873
TTFT (ms)
19.3
Peak mem (GB)
Hardware
Chip M4 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.5
macOS macOS 26.4
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,335 109.2 18.2 GB view
4k 1,681 102.0 19.0 GB view
8k 1,681 99.1 19.3 GB current
Batching Results
Batch Size TG tok/s Speedup
109.2 1.00×
165.2 1.51×
213.0 1.95×