Perplexity Introduces the Orchestrator API to Pick Models and Providers by Availability, Capacity, and Latency
Introducing the Perplexity Orchestrator API. Orchestrator lets you choose the right model for the right task, and automatically selects the provider based on availability, capacity, and latency. All open-weight and proprietary models are in one API with unified billing.
Context
Developers get one API with unified billing that chooses the model and provider for them, which could reduce outages and manual routing work.
Perplexity posted that the Orchestrator API lets developers choose the right model for the right task and automatically selects the provider based on availability, capacity, and latency. The post says all open weight and proprietary models are available in one API with unified billing. A link in the post leads to more detail.
Deep dive
Perplexity announced the Perplexity Orchestrator API. According to the post, Orchestrator lets developers choose the right model for the right task, and it automatically selects the provider based on availability, capacity, and latency. The post also says all open weight and proprietary models are available in one API with unified billing.
The post does not list which models or providers are included, pricing, rate limits, or how the automatic selection decides between providers. It includes a short video and a link, but only the post text was reviewed here, so those details are not described.
The design matches a pattern some developers use today, where they wrap several model vendors behind a single interface so that a busy or slow provider can be swapped out. Perplexity's pitch is that the choice of provider is handled for the developer while the developer still picks the model.
The announcement follows other Perplexity news in this deck, including its Frontier Research effort on Nov 16 and a post on annual recurring revenue crossing $1 billion on Nov 17.
What to expect next
Watch for the documentation, the list of supported models, pricing, and early developer reports on reliability and latency. Competition from other model routers and from the model labs' own APIs will determine whether the product becomes a default for developers. Perplexity's earlier posts this week on Frontier Research and revenue suggest the company is pushing hard on developer and enterprise products, so more API announcements may follow soon.
