Pricing
Prepaid credits with a single 3% platform fee at top-up. Usage is billed at the exact provider cost across 10,000+ models. Cache hits are a flat $0.25 per million tokens.
Top up any amount
Add any amount from $5 to $10,000. Your balance is prepaid, so there is never a surprise invoice at the end of the month. Auto-recharge is available if you want it, and it is off by default.
One 3% platform fee
Xantly's entire platform fee is 3%, charged once at top-up and shown at checkout. Pay $103 and receive $100 in credits. That is the whole fee.
Usage at exact provider cost
Usage draws down at the exact per-token price the upstream provider charges. There is no markup on usage: $100 of credits buys exactly $100 of provider usage.
| Line item | Amount |
|---|---|
| You pay | $103.00 |
| Platform fee (3%, shown at checkout) | $3.00 |
| Credits added to your balance | $100.00 |
| Provider usage those credits buy | $100.00 |
Cache and memory hits
Responses served from cache or memory cost a flat $0.25 per million tokens instead of the provider price. On workloads with repeated prompts this is where most of the savings come from.
What every account gets
- All 10,000+ models through one API, at the provider's own per-token rate. No model gates and no premium class held back.
- No member cap and no per-seat charge. A team draws down one shared credit balance.
- Bring your own provider keys and cloud credentials and route through them instead of ours. Nothing gates it.
- Monthly budget caps, threshold alerts and per-tool context masking, configurable from your dashboard.
- Single sign-on over OIDC with SCIM provisioning, set up by your own org admin. SAML connections are provisioned by Xantly on request.
- A hash-chained, tamper-evident audit trail of privileged actions.
Enterprise
Committed spend, invoicing instead of card top-ups, and a negotiated rate card bound to your organization. Signed data-processing terms, a contracted model allowlist on certified platforms with no cache or memory retention and pinned regions, and written region constraints matched against what the routing layer enforces. Availability terms are agreed case by case, and we do not publish a number we cannot yet measure.
Common questions
Is there a subscription or a free plan? No. Xantly has no subscriptions, no tiers and no per-seat charges. Everyone gets the same price: top up any amount from $5 to $10,000 and pay only for what you use.
Which models do I get? All of them. Every account gets the full catalog of 10,000+ models through one API, at the provider's own per-token rate.
How do cache hits save money? A cached or memory-served response is billed at a flat $0.25 per million tokens rather than the model's own price.
Do I have to talk to sales? No. Account creation, API key generation and top-up are all self-serve.