ORYVOX Core
One OpenAI-compatible API for every major model, with routing, caching, token optimisation, fallbacks, budgets and a full request log.
ORYVOX Core is the AI gateway every ORYVOX chatbot and agent already runs on, and you can use it for your own apps too. One API key and one endpoint give you every major model. Each request goes to the cheapest model that is good enough, reuses what was already paid for and sends fewer tokens.
If your code already uses the OpenAI SDK, point it at ORYVOX and you're done:
export OPENAI_BASE_URL=https://api.oryvox.ai/v1
export OPENAI_API_KEY=ory_live_...What you get
| Smart routing | Each request goes to the cheapest model that meets its quality floor. Choose Economy, Balanced, Quality or Max per project or per request. |
| Caching | Exact answers, answers to questions that mean the same, and the providers' prompt cache. Cached answers are free. |
| Token optimisation | Long conversations, pasted pages and tool results are trimmed before they reach the model. |
| Fallbacks | If a provider fails or is slow, the request moves to the next model without your code noticing. |
| Budgets | Per project, per day or month: block, downgrade to economy routing near the limit, or only warn. |
| Observability | Every request with its routing decision, cost and saving; traces, metrics, incidents and alerts. |
| Security | Key scopes, IP allowlists, zero retention, personal-data masking, audit log and SSO. |
Start here
Pricing in one line
Your Core plan covers the platform. Usage on ORYVOX keys is paid from credit: the provider's cost plus a small margin. Failed requests and cached answers are free. On your own provider keys (BYOK) there is no percentage. See Billing, budgets and idempotency.