← Back to community benchmarks

Qwen3.5-35B-A3B

M4 (10c) · 32 GB · 4bit · 2026-03-10
Performance
8k
tokens
440.3
PP tok/s
42.1
TG tok/s
18605
TTFT (ms)
20.4
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 32 GB
GPU Cores 10
Software
oMLX v0.2.6
macOS macOS 26.3.1
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 446.1 48.4 20.0 GB view
4k 458.5 45.9 20.2 GB view
8k 440.3 42.1 20.4 GB current
16k 393.3 35.0 20.8 GB view
Batching Results
Batch Size TG tok/s Speedup
48.4 1.00×
74.4 1.54×
86.2 1.78×