Perplexity Sonar
Sonar is Perplexity's lightweight, cost-effective search model optimised for quick, cited answers, with three search modes (High/Medium/Low).
Perplexity
Perplexity Sonar
Lightweight AI answer engine with real-time web search
Sonar is Perplexity's lightweight, cost-effective search model optimised for quick, cited answers, with three search modes (High/Medium/Low).
Key features
Perplexity Sonar · AI Models
Context window
128K tokens
Released
2026
Pricing
$1 / 1M tokens (input + output)
Key features
- Real-time web search with inline citations.
- Three search modes: High, Medium, Low.
- 128K-token context window.
- $1 per 1M input/output tokens.
Best for
Pick Sonar for cheap, fast, cited answers to factual queries — chatbots, research helpers, and any UX that benefits from grounded responses.
Frequently Asked Questions
Does Perplexity Sonar show citations?
Yes — every Sonar answer includes inline citations linking to the specific web sources used, so you can instantly verify any claim in the response.
What are the search depth modes in Sonar?
Sonar offers High, Medium and Low search depth modes. High searches more sources for better accuracy; Low is faster and cheaper for simple factual queries.
Open Chat
Sonar is Perplexity's lightweight, cost-effective search model optimised for quick, cited answers, with three search modes (High/Medium/Low).