| Input $/MTok | Output $/MTok | Best For | |
|---|---|---|---|
| Sonar (small) | $1 | $1 | Fast factual lookups, FAQ bots, high volume queries |
| Sonar Pro (large) | $3 | $15 | Comprehensive research, detailed answers |
| Sonar Reasoning | $3 | $15 | Multi step analysis with search grounding |
| Sonar Reasoning Pro | $5 | $25 | Complex research requiring deep reasoning and synthesis |
Migrating from OpenAI or Anthropic APIs
The Sonar API uses an OpenAI compatible format, so if you already have an application built on the OpenAI SDK, switching to Perplexity requires only changing the base URL and API key. The key advantage of Sonar over standard LLM APIs is that every response includes real time web search results and citations, so you do not need to build your own search and retrieval pipeline.
What is a token?
Tokens are the units AI models use to measure text. One million tokens (MTok) is roughly 750,000 words, or about 1,500 pages of text. "Input" tokens are what you send to the API (your question and context), and "output" tokens are the AI's response. Most simple queries use a few hundred tokens in each direction.