# Recruiting Credits budget: method and limits

Metix editorial design, official public pricing and Terms of Service checked 2026-10-30. This is a local planning tool, not a live product connector, invoice, quote or hiring forecast. No network requests are made.

Download all files together or unzip the kit. Python 3 standard library only:

    python3 recruiting-credits-budget.py recruiting-credits-examples.csv

Replace template values and blanks with your own counts. One row is one independent scenario for one actual subscription billing period. period_label is a user label, not a validated calendar. Do not split the same account's billable actions into multiple independent rows to reuse the allowance. Do not combine different accounts' allowances. These examples are invented; repeated period labels are alternative scenarios, never totals to add.

Standalone email unlocks use 10 Credits, phone unlocks 20, and Deep Search lists 100. Count newly charged unlocks/lists rather than repeatedly viewing already unlocked data. New hiring workflows beyond the monthly included AI Agent quota use 3,000 Credits each. Included paid plan (Credits; monthly new workflows): Starter (5,000; 1), Growth (10,500; 2), Professional (16,000; 3), Scale (27,000; 5). Extra Agents draw from the same Credit allowance, not a second free pool.

credits_required = 10 * standalone_emails + 20 * standalone_phones + 100 * standalone_lists + 3000 * max(0, new_workflows - included_agents).

active_workflow_actions records work already inside an eligible active Agent workflow. Its coefficient is zero. Classification is the user's responsibility: a standalone Deep Search list is not free simply because another workflow exists. The count does not grant unlimited usage, establish fair-use limits or predict results. It keeps workflow activity out of the standalone meter.

Unused monthly AI Agent quota does not roll over; closing a workflow does not restore it. Paid Credits expire at the next monthly anniversary, including on annual billing. Every row starts from that period's published allowance, with no invented carryover or pooling. Existing workflows do not consume a new-workflow unit just because the next month starts. The one-time trial is not a recurring paid-plan allowance and is not modeled.

Blank counts remain unknown; zero means confirmed no such actions. A missing count keeps total demand/headroom unknown, not zero. Within_allowance means only that entered modeled consumption fits. Allowance_gap shows missing Credits; it does not buy a top-up, permit automatic overdraft, approve an upgrade or assign a dollar/Credit price. Current Terms schedule plan changes for the end of the current billing cycle, so a larger plan is an alternative scenario, not an immediate fix. Confirm any additional capacity or custom terms with Metix before relying on it.

Optional internal labor is operator_hours * loaded_hourly_usd, rounded to cents. It is your assumption, not Metix staff time, an incremental cash invoice, a vendor saving, or an amount added to subscription cash. Blank labor inputs return unknown. The tool makes no currency conversions, plan-price forecasts, annual invoices, tax estimates, outcome/candidate/hire conversions, trial allocations or Delivery quotes. Subscription charges still apply when usage is zero. Self-serve and separately quoted staffed Delivery are different purchases.

Counts must be nonnegative whole numbers; hours/rate use plain nonnegative numbers with up to two decimals. A worksheet guard rejects inputs above 1,000,000; that is not a product limit. CSV headers must match, case IDs must be unique, and malformed files fail before output. CSV output blanks denote unknown; status identifies incomplete, within_allowance or allowance_gap. Exit code 0 means processing succeeded, not that a budget was approved.

Source facts: https://metix.ai/pricing ; https://metix.ai/terms-of-service sections 5-7 ; https://metix.ai/faq/credits . Recheck these before purchase or after a price/terms change.