April 2026 Summaries

Exa is a cutting-edge AI search engine offering tools such as a SERP API, website crawler, and deep research API, designed to enhance applications with web search capabilities. Recently, Exa partnered with Google, integrating its services into Google's Vertex AI and Gemini Enterprise platforms. The collaboration features "Grounding with Exa," which anchors Gemini's responses using fresh web data curated by Exa, ensuring accuracy and relevance in dynamic contexts like news or product data. Users can easily integrate Exa into their workflows within Gemini Enterprise via the Agent Marketplace, facilitating seamless web research and data retrieval. This partnership underscores Exa's readiness for implementation within Google Cloud infrastructures, aiming to improve user experience by providing timely and precise information retrieval.

Exa's search engine complexity has increased due to diverse user needs and the rise of AI agents, prompting the development of Canon, a search pipeline orchestrator. Canon represents search processes as a Directed Acyclic Graph (DAG), allowing for automatic parallelism, durable execution, and enhanced introspectability by decomposing the search process into manageable nodes. This structure facilitates efficient query handling by parallelizing independent nodes, allowing for concurrency, caching, and error tracing, which streamlines debugging and ensures reliability amidst growing demands. The system is designed to accommodate the predominance of code written by agents, turning implicit assumptions into explicit constructs, and leveraging a robust type system to ensure correctness. Canon's runtime efficiently manages orchestration, handling tasks like memoization and cancellation, while providing comprehensive observability that records every decision made during a query process. This approach supports Exa's need to deliver complex search capabilities across diverse customer requirements while maintaining quality and correctness at scale.

Exa has launched Deep Max, a cutting-edge search endpoint designed to tackle complex research questions with unprecedented speed and accuracy. By leveraging advanced large language models (LLMs) and making multiple parallel calls to Exa Search, Deep Max outperforms other agentic search systems and frontier LLMs in terms of accuracy and latency, completing queries significantly faster than competitors. This efficiency is achieved through parallel tool calls, token-efficient content processing, and Exa's rapid in-house search capabilities. As AI-driven web searches become more prevalent, Deep Max aims to provide fast, accurate, and transparent search results, positioning itself as the most sophisticated search solution available.

Exa, a modern AI search engine company, has opened its first Asia office in Singapore, focusing on core engineering for its retrieval stack to tap into the region's rapidly growing digital markets. Singapore was chosen due to its status as a hub for major tech firms and AI startups, its world-class universities, and strong governmental support for AI initiatives. This new office aims to enhance Exa's presence in the Asia-Pacific region, facilitating local time zone support for customers and offering engineers opportunities to work on next-generation search problems. The Singapore Economic Development Board has welcomed Exa's arrival, viewing it as a boon to the local AI ecosystem. With existing offices in San Francisco and Zurich, Exa is expanding globally, continuing its mission to build a web-scale infrastructure suited for AI, while actively recruiting engineers for various roles.

Exa is a cutting-edge AI search engine that offers a range of tools, including a SERP API and website crawler, aimed at enhancing web search and data retrieval processes. A key feature of Exa is its Highlights method, which extracts high-quality, relevant excerpts from web pages, significantly reducing token usage and improving accuracy in agentic search scenarios. These improvements allow Exa's engine to outperform competitors in terms of latency, cost, and quality, particularly in use cases that require multiple search rounds and efficient context management. The Highlights model is the result of extensive research, optimized for handling complex documents like coding and API references by delivering precise information from extensive text. Exa's agentic endpoints leverage Highlights for efficient retrieval, ensuring superior performance in various applications, with ongoing developments planned to further enhance the system's capabilities.

Exa has open-sourced WebCode, a new set of coding evaluations designed to improve the precision of web searches for coding agents, particularly in the context of code search which has seen a significant increase in queries. The initiative addresses the inadequacies of public benchmarks, which often fail due to issues like saturation and contamination, leading to unreliable evaluations of models' real-world programming capabilities. WebCode focuses on content and retrieval quality, evaluating the accuracy and completeness of web page content extraction, and retrieval precision, including citation precision and groundedness in search results. The evaluations are carried out using a dataset of 250 URLs and 317 query-answer pairs, employing a detailed methodology that ensures the extracted content is maximally useful for large language models in answering coding queries. The project aims to advance the industry standard for code search by providing a robust framework for evaluating both content extraction and retrieval processes, ultimately improving the performance and reliability of coding agents.