Price AI Spend the Way Your Contract Actually Reads

Govern access before spend happens

Provider list prices are a starting point, not the truth. Negotiated discounts, committed spend, credits, and cached-token rate differences all change what a customer or business unit actually pays — but bolting that onto a dashboard usually means a spreadsheet running in parallel with whatever the platform reports, and the two drift apart within a quarter.

• List pricing in the platform doesn't match the negotiated rate in the contract
• Commitments and credits tracked in a spreadsheet, disconnected from live usage
• No preview of what a rate change will do before it's approved and applied
• Historical reports break or silently change when a new rate is entered

One policy layer across model usage

Contract price sheets are customer/org-scoped and date-ranged, sitting alongside — never mutating — the global model catalog. Rate overrides can be set per provider, model, region, and date range, including cached-token and special-token rates. Before a price sheet or override is approved, Cloptima computes a server-authoritative preview against real historical usage, so you see the actual dollar effect before anyone signs off. Every price sheet, override, and approval is captured in an immutable audit history, and a commitment/credit ledger tracks draw-down separately from raw usage cost.

• Customer/org-scoped, date-ranged rate overrides — the global catalog is never mutated
• Cached-token and special-token rate rows, not just a flat per-token override
• Server-authoritative preview computed from real usage before approval, not a hand-built estimate
• Commitment and credit ledger with draw-down tracked separately from usage cost
• Immutable audit history behind every price sheet, override, and approval

Start with one app, then expand

Create a price sheet for your first contracted customer or business unit, add the rate overrides your contract specifies, run the preview against the last full billing period, and route it through approval before it goes live.

Built for private production AI

Effective-cost precedence is explicit and always explainable: contract override, then commitment, then credit, then retail fallback, then unknown-price handling if nothing else applies — so a historical report stays tied to the pricing rules that were in force at the time, even after new contracts are added later. Pricing changes rebuild a versioned overlay artifact rather than mutating the live catalog in place, so an approval never risks a half-applied rate change reaching production traffic mid-request.