Back to news

AI · Article

Perplexity Launches Orchestrator API With Same-Model Failover

Unite.AI ·

At a glance

Perplexity launched Orchestrator, an API that gives access to proprietary and open-weight models hosted on its own inference infrastructure. It offers unified billing, failover between deployments of the same model, and an optional automatic model routing feature in preview. Developers can keep using existing OpenAI or Anthropic SDKs by changing the API key, base URL, and model ID.

Key points

Why it matters

It can keep AI apps running when one provider fails, without changing models or per-token prices, which matters to developers who depend on these services.

Original source

Perplexity, 동일 모델 내 장애 조치를 지원하는 Orchestrator API 출시

Unite.AI · Nov 18, 05:42 KST

Read on Unite.AI ↗

AI summarized the original article. Check the original for accuracy. Copyright belongs to Unite.AI.

Perplexity · Ranking and more news