Plans, usage, and billing

Sample and Safe test use deterministic fixtures and do not spend model or browser allowance. Safe test runs a dry run. Run with my key uses your saved provider key in the active tab, so your model provider bills you directly. A hosted automation also uses its vaulted provider key when that route is available; otherwise it reserves Codelit's managed-model allowance. Browser Operator time is always metered separately when Codelit allocates the managed browser.

Open Settings > Security and data > Account to see today's my-key runs, this month's included model and browser use, active reservations, prepaid balances, hosted automations, and recent Execution Pack purchases. The run terminal, History, and Run Receipt keep the estimate, settled usage, and evidence for each completed run.

• Free includes 10 signed-in in-tab runs per day with one run at a time and 3 active Agent Apps.
• Pro includes $1 in managed-model allowance, 15 managed-browser minutes, 1 active hosted automation, and 25 active Agent Apps.
• Team includes 3 seats, $3 in managed-model allowance, 60 managed-browser minutes, 3 active hosted automations, and 25 active Agent Apps.

Before a managed Live Run, the launch review shows the projected model range, browser-time upper bound, and current balances. When the account has enough capacity, launch remains one click. Codelit opens funding choices only when the current plan or allowance cannot safely cover the reservation.

• Pro is $5/month for one builder and Team is $15/month for 3 seats when a Free account needs managed Live Runs.
• The $10 Execution Pack appears for an existing paid account only after settled usage leaves twenty percent or less of total capacity, or a specific Live Run projects a model or browser shortfall.
• Run with my model key keeps model charges with your provider when that route supports the Team.
• Test first runs the deterministic Sample without model or browser spend.

Codelit first reserves enough capacity to keep the run inside a hard spend boundary. After the stream finishes, model usage settles from provider-reported cost when available, provider token counts when cost is unavailable, or a clearly labeled token estimate as the fallback. Managed browser time settles to actual duration and returns unused reserved minutes.

• Provider cost or token settlement is labeled separately from an estimate.
• A pending receipt means the provider stream or browser session has not finished settling yet.
• Included allowance is consumed before prepaid Execution Pack balance.

An automation's daily run ceiling protects reliability and abuse boundaries; it is not prepaid execution. A run still stops when its managed-model or browser allowance is exhausted. Open-tab schedules use your active browser and keys, while hosted automations can continue after the tab closes.

After the contextual offer appears, Pro and Team owners can add a $10 Execution Pack from the blocked Live Run or Settings > Security and data > Account. Each pack adds $7 of managed-model credit and 300 managed-browser minutes after Stripe confirms payment. A run uses included allowance first, then prepaid balance.

• The payment session is verified against the signed-in owner, exact Stripe price, pack type, and quantity.
• Stripe webhook retries and the return-page verification share one purchase ID, so a replay cannot add the same pack twice.
• Prepaid balances carry forward and remain separate from monthly included allowance.

Included managed model and browser allowances reset monthly; prepaid balances carry forward. In-tab daily usage resets each UTC day. Owners can compare current plan limits on Pricing and open Stripe from Settings > Security and data > Account to view invoices, change payment, or cancel.