AI Agent News/stories/perplexity-launches-orchestrator-api-with-failover-and-model-routing.md

Infrastructure

Perplexity launches Orchestrator API with failover and model routing

Perplexity has launched an Orchestrator API that gives developers one endpoint for models from OpenAI, Anthropic, Google, xAI and open-weight providers. It adds same-model provider failover, unified billing and a preview router, perplexity/auto, that sends each request to a described route.

Source: www.marktechpost.com/2026/11/19/perplexity-orchestrator-api-failover-model-routing

The Orchestrator API serves models from OpenAI, Anthropic, Google, xAI and open-weight providers through a single endpoint. Billing for those models is unified in one place.

When a provider fails, the API can fail over to another provider that serves the same model. The routing layer is exposed as a preview router called perplexity/auto, which sends each request to a described route.

Because the router is still in preview, developers should expect its behavior to change before it is generally available. The announcement is brief, so details on supported parameters, pricing and route configuration should be checked directly with Perplexity.

Why it matters if you build agents

Agent builders can reach several model vendors through one endpoint and get same-model failover without writing their own provider-switching logic.

Since the auto router is in preview, teams should test which route it picks for their request types before relying on it in production.

Get stories like this every weekday

One short email with the day's agent tooling news.

Subscribe