Perplexity has introduced Sonar, a new API service designed to integrate its generative AI search tools into third-party applications. This move allows enterprises and developers to build real-time, citation-backed AI search into their platforms.
Perplexity highlighted that most AI features rely solely on training data, which limits accuracy. Sonar’s connection to trusted, real-time internet sources ensures its answers are both factual and reliable.
The Sonar API is available in two tiers: the base version, Sonar, which is affordable and fast, and the more advanced Sonar Pro, which handles complex queries with greater precision. Sonar’s base plan is priced at $5 per 1,000 searches and uses a lightweight model to provide quick and affordable AI search solutions.
Developers also pay $1 per 750,000 input words and another $1 for the same volume of output words. Sonar Pro offers deeper functionality, delivering more citations and detailed answers by running multiple searches per prompt. It costs $5 per 1,000 searches, $3 per 750,000 input words, and $15 per 750,000 output words.
Sonar’s affordability positions it as one of the most economical search APIs on the market. Already, companies like Zoom are leveraging Sonar to power AI assistants within their platforms.
Zoom’s integration allows users to access real-time, citation-backed answers directly in video chats without leaving the app.
Sonar is also an important revenue stream for Perplexity, which has an annual recurring revenue estimated between $5 million and $10 million. With recent price cuts in the AI tools market, Sonar aims to attract developers with its cost-effective pricing and functionality.
Perplexity raised $73.6 million this month, valuing the company at $520 million, signaling a push for further growth.
