# Perplexity API

> Perplexity provides a web-grounded API and SDK offering search-backed LLMs (Sonar family) and a Search API with token- and request-based pricing for developers building research and chat applications.

Perplexity provides an API and developer SDK that combine web search results with large language model reasoning to produce grounded answers and research outputs. The platform exposes a Search API billed per request and a set of Sonar models billed by tokens, including specialized Deep Research variants that return citation and reasoning tokens. Perplexity documents OpenAI-compatible chat completions endpoints and developer tooling (SDK, quickstarts, and examples) to integrate grounded LLMs into applications.

- **Search API (per-request billing)** — Use the Search API to retrieve raw web search results and filters; start by calling the Search endpoint and supply query and filter parameters. Pricing is request-based.
- **Sonar models (token pricing)** — The Sonar family (Sonar, Sonar Pro, Sonar Reasoning, Sonar Deep Research) charges by input/output/citation tokens; include tokens in your usage calculations and set model parameters in the SDK or API calls.
- **OpenAI-compatible chat completions** — The API supports a chat-completions style interface compatible with OpenAI formats; follow the chat-completions SDK quickstart to migrate or integrate.
- **Developer resources and community** — The docs provide quickstarts, API reference, a cookbook on GitHub, and a community forum and Discord for support. Start with the Perplexity SDK quickstart and create an API key in the dashboard.
- **Billing and admin controls** — The docs cover API groups, billing setup, rate limits, and key management so teams can monitor spend and control access.

## Features
- Web-grounded search blended with LLM responses (Sonar models)
- Search API billed per 1K requests
- Token-based pricing for Sonar model input/output/citation tokens
- OpenAI-compatible chat-completions endpoint
- Perplexity SDK quickstarts and API reference
- API key management, rate limits, and billing controls

## Integrations
GitHub (API Cookbook), Community Forum, Discord, AWS (hosting for models / infrastructure)

## Platforms
WEB, API, DEVELOPER_SDK

## Pricing
Paid

## Links
- Website: https://docs.perplexity.ai/guides/perplexity-sdk
- Documentation: https://docs.perplexity.ai/getting-started/overview
- Repository: https://github.com/ppl-ai/api-cookbook
- EveryDev.ai: https://www.everydev.ai/tools/perplexity-api