Billing & Usage

Cencori billing combines plan limits with usage-based credits.

1. Plan Tier: controls platform entitlements.
2. Credits Wallet: prepaid balance used for AI request charges.

Requests are not capped by count on any plan. Even with credits available, a request can still be blocked by:

• spend caps,
• or security/rate-limit controls.

Tiers differ by features and seats, not by a request allowance — no plan caps how many requests you can make. What you pay scales with what you actually use, via the credits wallet described below.

[!NOTE] Entitlements may vary by contract or deployment.

For each successful request, Cencori computes:

• provider cost from usage (tokens, image generations, audio duration/characters),
• applies configured markup,
• records final charge as cencori_charge_usd,
• deducts that amount from org credits.

Formula:

provider_cost + markup = cencori_charge_usd

Markup is model/provider-specific and configured in pricing data.

See Credits System for top-up and wallet details.

Use these controls to prevent runaway spend:

• Monthly Budget (soft): warning thresholds before overrun.
• Hard Spend Cap: blocks requests once cap is reached.
• Rate Limits: per-minute protection at project/API-key level.

See Bring Your Own Key (BYOK).

• Billing page (org-level balance and transactions)
• Project logs and observability views
• Request-level cost fields (provider_cost_usd, cencori_charge_usd, markup_percentage)