AI startup Perplexity has announced the “Perplexity Search API.” This will provide developers with infrastructure that enables the company to serve and an index that covers “thousands of billions” of web pages.
“When it comes to AI, the context is king. It's not enough to simply work at the document level. Index and search infrastructure divides the document into finer units,” Perplexity said in a blog post.
According to Perplexity, this search API is designed to meet the unique requirements of AI applications. “Unlike other API products that expose a limited universe of information, our APIs provide a wealth of structured responses that can be used in AI and traditional applications,” the company says.
Read: Rapid growth to $20 billion valuation of Perplexity hidden by Britannica legal lawsuit (
Perplexity said the new search API reduces the need for preprocessing, speeds up integration and provides more useful downstream results.
Price levels for this range range from sonar APIs to $1 input and output token per million, and $3 input and output tokens for Sonar Pro, respectively. It also offers specialized options such as Sonar Reasoning, Sonar Reasoning Pro, and Sonar Deep Research. This will vary in cost for inference, citations and search queries depending on the complexity of the workload. The company also says it has an advantage over competition in quality and delays.
Perplexity also has a search SDK, which engineers say could use it in AI coding tools to “develop impressive product prototypes within an hour.” “We expect even more impressive feats from startups, solo developers, mature companies and everyone in between,” the company added.
Read: Is Google Chrome on sale? Perplexity's $34.5 billion bid is a hot topic (
Perplexity recently reached a $20 billion valuation after winning fresh funds of $200 million. The company, led by Indian-American Aravind Srinivas, recently attracted attention with a $34.5 billion bid on Google's Chrome.
The startup is also said to have developed integrations with educational platforms and enterprise knowledge systems, establishing itself as the leading search solution for both work and personal use. But not everything is going well. Confusion, like many other tech companies, has recently faced accusations of copyright violations. Copyright holders, including the Encyclopedia Britannica and Merriam Webster, allege that the startup improperly used content on “Answer Engine” for online searches.
