Midium

Dedicated capacity

Hardware reserved for your team. One flat rate.

You don't buy, rack, or manage hardware. Subscribe for a flat rate, from $444/mo. We own, manage, and host it in our facility, with backup power, redundant network connectivity, and spare hardware on site. The machine is reserved for your team, and the bill does not move with tokens.

What you need it for

What do you want most? Required
What kind of AI do you need? Required

Select any that apply.

Tiers

Each tier lists a throughput range, a team size, and an agent capacity. A tier can be used for product traffic, larger agents, coding agents, or a mix of those.

The machines are Macs today, because that hardware is the best fit right now. A tier is what you get: the team it fits, the volume it holds, and the models it can run.

Pro

Throughput
3–5 billion tokens a month
Team size
11–20 engineers
Agent capacity
Frontier

Runs Laguna S on dedicated hardware.

Runs on a Mac Studio 96 GB

Dedicated capacity · $1,389/mo

Hub license · $500/mo

Max

Throughput
9–14 billion tokens a month
Team size
20–30 engineers
Agent capacity
Frontier

Includes Qwen3.8 Flash Next.

Runs on a Mac Studio 256 GB

Dedicated capacity · $2,149/mo

Hub license · $500/mo

Ultra

Coming soon

Throughput
22–28 billion tokens a month
Team size
30 or more engineers
Agent capacity
Frontier+

Includes GLM 5.3 Flash.

Runs on a Mac Studio 512 GB

Dedicated capacity · Coming soon

Hub license · $500/mo