
Perplexity Sonar Pro
Search-grounded API model: real-time web answers with citations, built for factual QA.
About
Launched January 21, 2025, Sonar Pro is Perplexity's premium search API model: instead of answering from frozen training data, it runs live web searches and grounds responses in retrieved sources with citations. Perplexity reported it beating Google, OpenAI, and Anthropic offerings on the SimpleQA factuality benchmark at launch, with roughly double the citation density of the base Sonar tier and multi-search behavior for complex, multi-step questions. Zoom integrated it into AI Companion 2.0 as an early enterprise customer.\n\nThe honest caveats: pricing is search-plus-token based ($5 per 1,000 searches plus $3/1M input and $15/1M output tokens), and because Pro runs multiple searches per prompt, costs can be unpredictable. Real-time grounding helps factuality but doesn't eliminate hallucinations, and outputs are non-deterministic — a poor fit for applications needing exact reproducibility.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 51; Trending score 74/100 from researcher assessment.