Perplexity has unveiled its new Search API, designed to give developers access to the same global-scale infrastructure that powers its public answer engine. The API draws from an index of hundreds of billions of webpages, making it one of the largest developer-accessible search offerings on the market.

Alongside the API, Perplexity is also releasing a Search SDK, an API Platform with developer console and documentation, and an open-source evaluation framework called search_evals, plus a deep dive from Perplexity Research into the design and evaluation of its search technology.

Unlike traditional APIs limited to document-level results, the Perplexity Search API is built for AI-first applications. Its infrastructure divides documents into sub-document units, which are individually ranked against query parameters. This allows the API to return results with highly relevant snippets already surfaced, reducing preprocessing and enabling faster integration.

“When it comes to AI, context is king,” the company emphasized, noting that this granular approach supports more valuable downstream results for both AI and traditional applications.

Since its founding, Perplexity has focused on accuracy and trust, pioneering source corroboration in AI answers. The new Search API builds on that foundation, using indexing workflows optimized for content freshness and realtime accuracy.

Perplexity’s systems process tens of thousands of index update requests per second, ensuring results remain up to date. An AI-powered content understanding module dynamically generates parsing logic to handle the variability of the open web, continuously refining itself through iterative AI self-improvement informed by millions of user queries each hour.

These systems, combined with robust evaluations and realtime signals, position Perplexity as a leader in accuracy and relevance. Developers can now leverage this infrastructure across industries ranging from enterprise knowledge tools to next-generation AI agents.

The Perplexity Search API aims to balance speed, quality, and cost. The company is releasing an open-source evaluation framework called search_evals, enabling developers and researchers to benchmark search APIs. According to Perplexity, the API outperforms competitors across both single-step queries and deep research workflows, all with lower costs.

Beyond performance, the API was designed for usability. The API Platform now houses the developer console and documentation for both the Search and Sonar APIs, centralizing resources for developers, while the Search SDK allows rapid prototyping. Perplexity engineers have reportedly built product prototypes in under an hour using the SDK, demonstrating its ease of integration.

In an effort to democratize access, Perplexity is offering developers a chance to get started on the API Platform, install their SDK, and read more about their research.

The company will showcase the Search API at its upcoming San Francisco API Day and London hackathon next month, and developers are encouraged to join the community forums for ongoing support.

Q: What is the Perplexity Search API?
A: The Perplexity Search API gives developers access to the same infrastructure and index of hundreds of billions of webpages that power the Perplexity public answer engine.

Q: How is it built for AI workloads?
A: The API surfaces fine-grained document snippets ranked for relevance, reducing preprocessing and improving AI application performance.

Q: How does it ensure accuracy and freshness?
A: The system processes tens of thousands of index updates per second, uses AI-powered parsing, and incorporates realtime signals from millions of queries.

Q: What tools are available for developers?
A: Perplexity is releasing a Search SDK, a centralized API Platform, and an open-source evaluation framework called search_evals.

Q: What’s next for the Search API?
A: Perplexity will continue improving performance and cost-effectiveness while showcasing the API at developer events like API Day and hackathons.

The launch of the Perplexity Search API marks a significant moment in the evolution of developer tools for the AI ecosystem. By opening access to its internet-scale infrastructure, Perplexity is addressing one of the key challenges developers face: reliable, accurate, and fresh retrieval at scale.

For developers, this could reduce the barriers to building AI agents, enterprise applications, and consumer tools that depend on real-time knowledge. For the market, it underscores a broader trend of search incumbents being challenged by newer players offering openness and developer focus.

If adoption accelerates, Perplexity may position itself not only as an AI answer engine but also as a foundational layer of the AI-driven internet, where both humans and code can access knowledge at scale.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.