Transparent pricing. No surprise costs.

You pay the real token cost of each request, exactly as billed by the AI provider, plus a clear platform fee (8%, 3%, or 2% depending on plan). We never mark up tokens silently. The fee is shown on every invoice.

On the Team plan and above you can connect your own provider credential — a Claude subscription token, an Anthropic API key, an OpenAI API key, or an OpenRouter API key — and route your team’s Claude Code, Codex, and OpenCode traffic through it. You pay the provider directly, so Clawgate never charges you for those tokens; we only apply the platform fee on the calculated usage. For OpenRouter, Clawgate sends "data_collection": "deny" with every request, so it skips any upstream provider that may train on your inputs. You keep unified budgets, per-user analytics, and abuse monitoring across every key, whether it runs on metered inference or your own provider credential.

Paid plans bill a fixed monthly fee per seat plus your metered usage and platform fee — added together once, never double-counted. The seat fee renews once per monthly billing cycle that starts the day you sign up or switch plans (not the 1st of the calendar month). Usage is billed at the end of each cycle, or sooner if it crosses a threshold you set. Your dashboard always shows a live month-to-date breakdown and projection, so there are no surprises.

You keep your current plan’s features until the end of the cycle you’ve already paid for, then your account switches to the cheaper plan automatically — and nothing extra is charged when you schedule it. For example, if your cycle renews on the 15th and you move from Team to Pay As You Go on the 5th, you keep every Team feature through the 15th, and Pay As You Go takes over from the next renewal. Your dashboard shows the exact date the switch takes effect.

No. As long as you switch back to your current plan before the scheduled date, the pending downgrade is simply cancelled. There’s no new per-seat charge and your billing cycle isn’t reset — you already paid for this cycle, so you carry on exactly as before. For example, downgrade Team → Pay As You Go on the 5th, then pick Team again on the 10th, and the switch just disappears at no cost.

Upgrades take effect immediately and are prorated, so you only pay the difference — your renewal date doesn’t change. We credit the unused part of your current seat fee and charge the prorated cost of the new plan for the rest of the cycle. For example, a 20-seat account half-way through the month moving from Team ($5/seat) to Business ($10/seat) is credited the unused ~$50 of Team and charged ~$100 of prorated Business, so you pay about $50 now; Business then renews in full on your normal renewal date.

The scheduled downgrade is dropped and your upgrade applies right away, with the same prorated billing as any other upgrade. For instance, if you’re on Team with a switch to Pay As You Go pending and you move up to Business, the Pay As You Go switch is cancelled and you go straight to Business — paying only the prorated difference for the rest of the cycle.

Pay As You Go has no seat fee at all. Team is billed for a minimum of 6 seats and Business for a minimum of 20, even if you have fewer active users — the seat fee is the larger of your active user count and the plan minimum. Above that floor you’re only billed for the seats you actually use.

Budgets are checked before each request runs. When a user or project hits its daily or weekly token, session, or dollar cap, the request is blocked. Spend simply cannot go over the limit you set.

No. Failed requests are never billed, and prompt-cache savings are passed straight through to you, never marked up.

Enterprise runs in your own infrastructure, so you pay your AI provider directly for tokens. You pay an annual platform licence instead of a per-token markup, and your data never leaves your environment.