← Back to community benchmarks

Qwen3.5-0.8B-Claude-4.6-Opus-Reasoning-Distilled

M4 (10c) · 16 GB · 8bit · 2026-04-29
Performance
16k
tokens
1,853
PP tok/s
83.6
TG tok/s
8841
TTFT (ms)
2.9
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 16 GB
GPU Cores 10
Software
oMLX v0.3.8rc1
macOS macOS 26.4.1
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,522 107.7 1.7 GB view
4k 1,972 100.1 1.9 GB view
8k 1,974 95.4 2.3 GB view
16k 1,853 83.6 2.9 GB current
32k 1,510 60.1 4.1 GB view
Batching Results
Batch Size TG tok/s Speedup
107.7 1.00×
176.6 1.64×
299.7 2.78×
324.9 3.02×