Key Takeaways

• A hybrid pricing model pairs a fixed component with a variable one, so one invoice carries both a subscription line and a usage line.

• Charges resolve in sequence: allowance, rating, discounts, then credits. Swapping the last two changes the total.

• A wallet scoped to usage charges gives a different total than one applied to the subtotal. On the invoice below, the gap is $160 on a $565 subtotal.

• Mid-cycle plan changes are hybrid's own failure mode: the fee prorates and the allowance has to prorate at once, while usage keeps accruing.

What actually combines in a hybrid pricing model, and in what order do the charges resolve?

Four parts combine, and they resolve in sequence rather than all at once.

• Fixed component. A platform fee, plan fee, or per-seat charges, billed in advance and known on day one.

• Included allowance. A quantity the fixed fee already covers, consumed before any charge exists.

• Variable component. Metered usage above the allowance, rated per unit or through tiers. That's consumption-based pricing running as one part of a larger charge.

• Balance component. A prepaid wallet, promotional credits, or a commitment drawdown, applied after rating.

The order that produces a defensible invoice runs allowance, rating, discounts, then credits. Credits come last because a percentage discount applied after a drawdown cuts money the customer already paid for. When the components collapse into one line, what you get is a composite charge.

How do a base fee, a metered overage, and a credit drawdown add up to one invoice total?

Each line resolves against its own basis first, then the wallet applies only to the lines it's scoped to.

The wallet keeps $160.00 because the restriction held. An engine that ignores scope draws the full $250.00, bills $315.00, and empties the wallet: a $160.00 difference on identical usage. That sequence is what credit burn-down governs, and getting it wrong stays quiet: the arithmetic still balances.

Why does a mid-cycle plan change get harder when usage is already accruing?

Two clocks prorate at once and only one has a settled answer. A pure subscription prorates one fee. A pure usage model has nothing to prorate. Hybrid prorates the fee and the allowance together, mid-period, against usage already consumed under the old allowance.

Take the same account moving to a $900 base on day 18 of a 30-day cycle, allowance rising to 6,000,000 tokens, usage 9,000,000 tokens split 1,800,000 before and 7,200,000 after.

• Prorated fee, the settled part: ($400 x 17/30) + ($900 x 13/30) = $616.67.

• Allowance resets in full to 6,000,000: overage 3,000,000, usage line $27.00.

• Allowance prorates to 3,733,333: overage 5,266,667, usage line $47.40.

• New allowance plus 200,000 unused, 6,200,000: overage 2,800,000, usage line $25.20.

All three are defensible, and the spread is 88% on the usage line for identical consumption. We make it a policy the billing cycle configuration states outright, before a customer asks why an upgrade raised their overage.

What has to be true in your billing stack before hybrid pricing is safe to run?

Five capabilities have to exist, and missing any one of them pushes the arithmetic back into month-end spreadsheets.

• Subscription, metered, and one-time charges sharing one period and one invoice.

• Allowances modelled as entitlements the meter draws against, not plan-page text.

• Credit wallets with scope, priority order, and expiry, applied after discounts.

• Proration policy for the fee and the allowance, set per plan and versioned.

• Invoice preview against live usage, and a per-line audit behind each amount.

Flexprice is enterprise-grade, open source usage based billing infrastructure for AI and SaaS companies. It can be deployed in your own VPC, on-prem, or on Flexprice's managed cloud. It puts subscription, metered, and one-time charges on the same invoice, and the engine applies the credit drawdown and proration rather than a person at month-end. As Navendu A., Head of Business, put it: "Our pricing changes every time we ship a new model, and that's a lot. Flexprice is the only tool that's kept up." Both halves get configured together on Pricing Models.

Related terms

Hybrid puts several billing mechanisms in the same period, so the terms below each name one part that has to agree with the others.

• Composite Charge is the billing object a hybrid model produces when its components collapse into one line.

• Credit Burn-Down sets the order and scope in which wallet balances get consumed.

• Consumption-Based Pricing is the variable half of a hybrid model standing on its own.

• Billing Cycle defines the period every hybrid component has to resolve inside.

• Consolidated Invoicing is what happens when hybrid charges from several accounts roll onto one document.

• AI Pricing Models covers the token, run, and outcome structures that usually sit in the variable slot.

FAQ

Is a hybrid pricing model the same as tiered pricing?

No. Tiered pricing bands one dimension into steps, such as $0.90 per 100,000 tokens up to 5 million and $0.70 above it. A hybrid pricing model combines two mechanisms, a fixed charge and a variable one, and the variable half is very often tiered internally.

How is a hybrid pricing model different from consumption-based pricing?

Consumption-based pricing charges only for what a customer uses, so revenue drops to zero in a month with no activity. A hybrid pricing model keeps a fixed floor under that, and the invoice is the tell: consumption-based produces usage lines only, hybrid produces both.

Should credits apply before or after a discount?

After. A discount reduces what's owed and credits settle what's owed, so drawing the wallet first means the discount cuts money the customer already prepaid. Rate the usage, apply discounts, then draw the wallet against what remains.