← Back to community benchmarks

Qwen3.5-2B

M3 Ultra (60c) · 256 GB · bf16 · 2026-03-12
Performance
32k
tokens
4,082
PP tok/s
99.7
TG tok/s
8028
TTFT (ms)
6.2
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 256 GB
GPU Cores 60
Software
oMLX v0.2.10
macOS macOS 26.3
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,243 135.4 4.9 GB view
4k 3,824 130.1 5.0 GB view
8k 4,730 124.0 5.1 GB view
16k 4,531 112.9 5.5 GB view
32k 4,082 99.7 6.2 GB current
Batching Results
Batch Size TG tok/s Speedup
135.4 1.00×
118.8 0.88×
233.6 1.73×
437.9 3.23×