Frequently Asked Questions

apicuts is a done-for-you API gateway for all models. You point your SDK's base URL at your apicuts endpoint, and your requests flow through us to the model providers you already use.

We analyze the traffic, find the highest-cost calls, and switch on the fixes inside the gateway: caching, model routing, batching, prompt reduction, and output limits.

Your stack stays the same. We then compare cost per equivalent request against your baseline.

You leave with:

• The exact call sites driving spend
• Your estimated monthly overpayment
• The highest-impact fixes
• A clear recommendation on what we can ship first

No generic advice. We use your actual usage data.

Yes. They should be able to.

The problem is usually not whether they can write the code. It is finding the waste quickly, prioritizing the highest-value fixes, and validating that savings do not damage quality or latency.

We run all of that for you inside a managed gateway, so your team keeps shipping product.

One working day from kickoff to live traffic. You swap one base URL and we take it from there.

Savings are then verified against your real usage over an agreed measurement period, typically 14 days.

Yes.

apicuts exposes an OpenAI-compatible endpoint, so switching is usually one base URL and one API key. Your prompts, tools and code stay the same.

Yes.

You get the model you ask for, with the same quality and capabilities. Cheaper equivalent models are used only where you approve them.

Only to verified, authorized upstream providers.

Each request goes to the cheapest suitable provider for the model you asked for.

Requests fail over automatically to a healthy route, so production keeps running.

Yes.

With BYOK (Bring Your Own Key), you keep your existing provider accounts.

No. A change that doesn't meet your quality and latency requirements doesn't ship.

We test against your existing evaluations, or help create a representative evaluation set before anything is switched on. We track cost, latency, error rate, and output quality before and after every change.

We define the baseline before go-live.

If your all-in cost per equivalent request, meaning provider charges plus our fee, isn't at least 20% lower, you pay us nothing. Charges you've already paid to model providers aren't refunded.

We don't store them by default.

Your requests pass through the gateway to your model providers, and prompts and outputs are not saved. We log only metadata: token counts, model, cost, and latency. If you switch on semantic caching for a route, its cached responses are kept only until they expire.

For the teardown, a usage export is enough. If you share logs, you can redact sensitive information first.

Not by default. No prompts, outputs, or user content are stored unless you switch on semantic caching for a route, and those cached responses expire.

We keep request metadata for your savings reports and delete it when you leave. Every vendor that touches your data is listed in our Trust Center.

In the EU or the US, your choice.

Routing keeps your requests with providers and regions that match.

Our fee is 10% of the savings we verify. No retainer and no long-term contract.

The 20% guarantee is measured after our fee: if your all-in bill isn't at least 20% lower, you pay nothing. See pricing.

Yes.

Set budgets, usage limits and alerts, so there are no unexpected bills.

Yes. Enterprise SLAs and support are available for production workloads, and a public status page with live uptime is coming soon.

No. No long-term contracts and no retainers.

No.

Leaving takes one line: point your base URL back at your provider. No contracts or retainers.