← Back to community benchmarks

Qwen3-VL-8B-Instruct-oQ8

M1 Ultra (48c) · 64 GB · 8bit · 2026-05-10
Performance
4k
tokens
760.4
PP tok/s
29.8
TG tok/s
5386
TTFT (ms)
10.6
Peak mem (GB)
Hardware
Chip M1 Ultra (48c)
Memory 64 GB
GPU Cores 48
Software
oMLX v0.3.8
macOS macOS 26.3
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 750.0 32.4 9.9 GB view
4k 760.4 29.8 10.6 GB current
Batching Results
Batch Size TG tok/s Speedup
32.4 1.00×
53.7 1.66×
72.7 2.24×