Perplexity announced an Orchestrator API that lets developers call one endpoint and have the request routed to models from OpenAI, Anthropic, Google, xAI or open‑weight providers. The API can automatically fall back to another model from the same provider if the first one fails. Billing is combined into a single invoice, and a preview auto‑router decides which route to use for each request.
Why it matters
Developers can simplify integration and improve reliability by using a single API instead of managing separate connections to each AI provider.