fast/ultrafast service tiers for private models

OpenAI/codex have fast/ultrafast tiers which offer higher inference speeds at a higher price.

I would love to have a similar option to pay more for much faster inference speeds for the private models on Venice via API.