Perplexity Begins Private Preview of Orchestrator API for Developers
The catalog includes models from Moonshot AI, Z.AI and DeepSeek. Developers must request preview access by email and use an existing Perplexity API key.

Perplexity began a private preview of its Orchestrator API on Nov. 9, giving developers a single gateway to access Perplexity-hosted open-weight models with one bill.
The initial catalog lists perplexity/kimi-k3 from Moonshot AI, perplexity/glm-5.3 and perplexity/glm-5.3-flash from Z.AI, and perplexity/deepseek-v4-pro-0813 from DeepSeek. Developers can retrieve the model catalog through a GET /orchestrator/v1/models request.
The API accepts request formats used by OpenAI and Anthropic, regardless of which company created a model. Developers use one base URL and an existing Perplexity API key.
The service monitors model deployments and can shift traffic to another deployment if one has a problem. Billing is pay-as-you-go at published per-token rates, listed per million tokens. Requests rejected for rate-limit errors carry no charge.
Developers must request private-preview access by email and authenticate with their existing Perplexity API key.