← Back to community benchmarks

Qwen3-VL-Embedding-2B

M4 (10c) · 16 GB · 4bit · 2026-03-10
Performance
8k
tokens
826.2
PP tok/s
43.9
TG tok/s
9917
TTFT (ms)
3.1
Peak mem (GB)
Hardware
Chip M4 (10c)
Memory 16 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 870.4 81.5 2.5 GB view
4k 925.8 59.4 2.7 GB view
8k 826.2 43.9 3.1 GB current
16k 670.7 29.4 3.9 GB view
Batching Results
Batch Size TG tok/s Speedup
81.5 1.00×
124.2 1.52×
156.8 1.92×