Pricing disclosure.

We are not a token reseller.

Synapse Garden is a developer-experience product built on top of a shared API. The list price you see for any model is calculated like this:

You don’t need to do this math anywhere — the catalog page shows the marked-up number directly. We only document it here for full transparency.

• Per-project API keysIssue, hash, rotate, revoke — without redeploys.
• Atomic spend capsHard caps per project return 402 once exceeded.
• Audit logEvery dashboard action, 90 days hot.
• Usage dashboardsDaily roll-ups within 5 minutes.
• Single SDK callOpenAI- and Anthropic-compatible base URL.
• Model allowlistsPer-project restrict which models you call.
• Failover routingProvider hiccups don’t take you down.
• SupportEmail response same business day during beta.

Each plan ships with a chunk of included tokens. When you cross that threshold, additional usage bills at the same list rate as the catalog. There is no penalty rate, no “burst” multiplier, no surge.

We don’t bill you for cache reads (provider-side prompt caching is honored), retried requests during a provider failover, or the first response when the upstream returns a 5xx. If you spot a discrepancy, email billing@synapse.garden and we’ll reconcile.