Perplexity Launches Sonar API, Promising Real-Time AI Search Solutions
Perplexity has launched its real-time AI-powered search tool API, Sonar. It is designed to provide accurate and relevant answers. Available in Sonar and Sonar Pro versions, it caters to basic and advanced query needs and aims to redefine AI search technology for businesses.
It is the first of its kind, as the company mentioned on its website, “While most generative AI features today have answers informed only by training data, this limits their capabilities. To optimize for factuality and authority, APIs require a real-time connection to the Internet, with answers informed by trusted sources.”
Base Version: Cost-Effective and Efficient
The base version of Sonar offers an economical way to add AI search to applications. It costs $5 for every 1,000 searches and uses a lightweight model that enables quick responses.
However, you will also have to pay $1 for every 750,000 words entered into the model (around 1 million input tokens) and $1 for every 750,000 words generated (about 1 million output tokens). This simple pricing will help businesses control costs while accessing reliable AI tools.
Sonar Pro: Advanced Capabilities for Complex Queries
On the other hand, Sonar Pro is designed to provide detailed answers to complex queries. Like the base version, it costs $5 per 1,000 searches but handles more in-depth processing. Pricing includes $3 for every 750,000 input words and $15 for the same number of output words. It also offers twice as many citations, making the retrieved information more reliable.
| Feature | Base Version | Sonar Pro |
| Target Use | Basic queries, cost-effective AI | Complex queries, detailed answers |
| Cost per 1,000 searches | $5 | $5 |
| Cost per 750,000 input words | $1 | $3 |
| Cost per 750,000 output words | $1 | $15 |
| Response Speed | Fast (lightweight model) | Advanced (multiple searches) |
| Number of Citations | Standard | Twice as many as the Base Version |
Performance Metrics That Impress
The company has also shared SimpleQA benchmarking results, showing that its latest Sonar Pro is topping the chart and leaving behind big players like Google, OpenAI, and Anthropic. The Sonar Pro achieved an F-score of 0.858, while Sonar received an F-score of 0.773.
Real-World Applications in Action
Tech companies have already started integrating Sonar into their AI backend. Here are some examples:
Copy AI
Copy AI has integrated Sonar into its platform to assist go-to-market teams in researching prospects and target companies. If we go by the Perplexity blog post, this integration has proven to be a game-changer, enabling companies to save an impressive 8 hours of research per representative each week, resulting in a 20% increase in throughput.
Doximity
Doximity, a platform dedicated to healthcare professionals, is also using Sonar to provide doctors with a robust research tool to elevate their practice. Thanks to Sonar’s research-backed answers, medical professionals can get information about new guideline changes from medical journals or clarify questions about insurance reimbursements. Given the high-stakes nature of medicine, including in-line citations is critical for ensuring accuracy and building trust among users.
Zoom
The popular video conferencing platform has also integrated Sonar Pro with its AI Companion 2.0, which incorporates various services directly into its platform. The integration of Sonar Pro allows users to ask any question without leaving the video call or opening a browser.
According to Will Siegelin, Zoom’s Product Manager of AI Products and responsible AI, “Perplex” opens Zoom to knowledge beyond its four walls.
That’s a wrap for now, but with Sonar’s API release, one thing is clear—in the coming time, we will all likely be using Perplexity’s AI search engine, directly or indirectly. Currently, no other company offers real-time AI search capabilities like Perplexity.
However, with industry giants like Google and OpenAI in the game, it will be interesting to see how Perplexity handles the competition when it arises.