← Back to community benchmarks

Qwen3.6-35B-A3B

M2 Ultra (60c) · 128 GB · 4bit · 2026-04-20
Performance
4k
tokens
1,130
PP tok/s
78.2
TG tok/s
3625
TTFT (ms)
20.0
Peak mem (GB)
Hardware
Chip M2 Ultra (60c)
Memory 128 GB
GPU Cores 60
Software
oMLX v0.3.6
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 868.6 83.2 19.3 GB view
4k 1,130 78.2 20.0 GB current
Batching Results
Batch Size TG tok/s Speedup
1× 83.2 1.00×
2× 136.2 1.64×
4× 194.5 2.34×
8× 254.6 3.06×