Pricing

Mate is priced per mate — one mate is one concurrent agent slot. A team of seven typically shares a single mate; buy more when one slot stops keeping up. The price is fixed and scales linearly: adding more mates costs the same per mate regardless of quantity.

For current prices and a plan comparison table, see the pricing section on the marketing site.

Plans at a glance

Shared — multi-tenant infrastructure in the EU. Jobs run in isolated containers on shared workers.

Dedicated — reserved compute assigned to your mate whenever a job runs. No idle billing.

On-prem — the Mate binary runs on your own hardware with a Postgres container alongside it. Unlimited concurrent slots, bounded only by your hardware. Billed annually.

LLM usage

Mate does not bill for LLM token usage. In the current model, you bring your own LLM API key (BYOK) and tokens are consumed directly against your provider account. Mate’s billing covers compute capacity only.

For details on connecting an API key, configuring per-job spend caps, and supported providers, see Models and BYOK.

Billing during beta

The console billing view is not yet active. During the closed beta, subscriptions are handled manually. If you have questions about billing, contact the team (see Support).

Prices stated on the marketing site are introductory and locked in before general availability. Beta participants keep beta pricing for 12 months from GA.

All prices are ex. VAT.