According to a recent LinkedIn post from Perplexity, the company is presenting Orchestrator, an API designed to route requests among more than 45 models from OpenAI, Anthropic, Google, xAI and open-weight providers. The post describes unified billing and provider selection based on availability, capacity and latency, positioning the service as an abstraction layer for developers using multiple model vendors.
The described API supports OpenAI Chat Completions, Responses and Anthropic Messages interfaces, allowing existing SDK connections to use a Perplexity API key. It also includes retrying failed requests on another deployment of the same model, along with controls to track spending by model, project and API key and set model allowlists and spending caps.
Perplexity’s post also references automatic routing in preview, with profiles translating plain-language tasks into model choices and a decision model selecting routes. It says the underlying model layer currently routes hundreds of millions of requests across more than 85 models and a dozen providers, although the post does not provide revenue, customer or performance data.
For investors, orchestration could broaden Perplexity’s role from an AI search and services provider to an infrastructure intermediary, potentially supporting API usage and recurring usage-based revenue. The financial impact will depend on adoption, pricing, provider costs, reliability and differentiation in a crowded model-routing market, while the multi-provider approach could increase operational complexity.


