One credit per model call. No seat tax, no agent caps you'll actually hit.
A credit reflects the model work a task actually uses — quick questions sip a few, deep reports draw hundreds.
Whatever model an agent runs, its usage lands in the same unit — no per-model token math to compare or forecast.
Claude, GPT, or Gemini per agent; the credit rate follows the model you pick.
On Business and Enterprise, pay your provider directly and credits become a thin processing fee.
Everything you need to know about the product and billing.
Bring a real workflow. We'll wire it live, ship it to your Slack by end of call.