March 2025 Summaries

We recently disclosed a critical vulnerability in Next.js, which was discovered through private GitHub reporting. The vulnerability affected older versions of Next.js (12.x) and impacted self-hosted applications using `next start` and standalone output. After investigating the report, our security team confirmed the validity of the issue and began researching remediation options. We released patches for Next.js 14.2.25 and 15.2.3, and also published a changelog on Vercel stating that customers were protected. The vulnerability was later made public by GitHub, but we had already communicated with affected partners and provided additional support. Our team is now improving how issues get reported, enhancing our process for responding to disclosures, and implementing new security measures to prevent similar vulnerabilities in the future.

The AI SDK is an open-source toolkit for building AI applications with JavaScript and TypeScript, providing a unified provider API that enables powerful UI integrations into leading web frameworks. The latest release introduces several key features, including Reasoning models, which dedicate computational resources to work through problems methodically, producing more accurate and reliable results. Additionally, the SDK now supports Model Context Protocol (MCP) clients, which connect applications to a growing ecosystem of tools and integrations, allowing developers to access hundreds of pre-built tools that add powerful functionality to their application. The AI SDK also introduces new features such as image generation with language models, message parts, URL sources, OpenAI Responses API, Svelte 5 support, and middleware updates. These features enable the creation of multi-modal chatbots, improved error handling, and enhanced provider options. With these new capabilities, developers can build more powerful and reliable AI applications, including agentic spreadsheets like Otto and full-stack frameworks like Payload. The SDK is maintained by a core team at Vercel and has received contributions from many community members, with special recognition given to individuals who have made significant contributions to the project.

Vercel is partnering with xAI to bring their powerful Grok models directly to Vercel projects through the Vercel Marketplace, enabling quick prototyping and experimentation without additional signup required. This integration simplifies authentication, API key management, and billing through automatically configured environment variables. Developers can access Grok's large language models (LLMs) directly from their Vercel projects, simplify authentication and API key management, pay only for what they use with integrated billing, and explore a wide range of AI model types, sizes, and specializations in the Vercel Marketplace. The partnership is part of Vercel's ongoing effort to make using AI frictionless on its platform.

Vercel is a company that aims to make building for the web more accessible and powerful, with a vision of empowering developers to bring their biggest ideas to life. The company has recently welcomed Jeanne DeWitt Grosser as its Chief Operating Officer, who will lead the go-to-market function, bringing strong leadership and operational excellence to further this mission. Vercel is building a foundation to power the next billion developers, with a focus on making AI more accessible and building a platform to meet the demands of modern web applications. The company's leader Jeanne embodies a "leader who ships" approach, knowing how to scale developer-first companies and has experience shaping go-to-market strategies at influential companies like Stripe and Google. Vercel is poised to shape the future of development with AI, making it an exciting time for the company as it grows under Jeanne's leadership while staying true to its core values.

The text discusses the importance of personalization in business, its two core types - architectural and content personalization, common mistakes to avoid, and how Next.js and Vercel provide tools for implementing personalization effectively. It highlights the benefits of personalization, including increased revenue, improved user experience, and reduced costs. The article also provides guidance on how to implement personalization incrementally with the right tools, without sacrificing performance or creating inconsistencies. Additionally, it mentions that personalization is crucial for ecommerce brands, as 65% of consumers expect brands to adapt to their individual needs and expectations, and teams that implement successful personalization are 48% more likely to exceed their revenue goals.

Sitecore, a leading digital experience platform, developed Sitecore Stream, a dynamic AI-powered brand assistant that empowers marketers to interact with their brand content in a visually interactive and conversational way. The AI SDK used by Sitecore provides a seamless integration of AI with front-end user experience, prioritizing both AI performance and user interaction. With the AI SDK, developers can build interactive, AI-driven applications with tools and components designed to simplify development workflows. Sitecore's brand assistant uses Python on the backend, Next.js on the frontend, and the AI SDK to bring them together, leveraging real-time streaming to deliver fast responses that feel immediate and engaging. The AI SDK enables marketers to ask questions, brainstorm ideas, and generate content in real-time, providing an interactive and conversational experience where they feel like they're working with a virtual creative partner that understands their brand. Sitecore envisions a future where their clients can use features like brainstorming or finding visual assets, each requiring unique user experiences, and has laid the foundation for a future where AI-powered brand interaction is intuitive and deeply integrated with the unique needs of each brand.

Vercel's Fluid compute is a next-generation compute model designed to handle modern workloads with real-time scaling, cost efficiency, and minimal overhead. It addresses the inefficiencies of traditional serverless architectures by dynamically adjusting to traffic demands, reusing existing resources before provisioning new ones. The Vercel Functions router orchestrates function execution to minimize cold starts, maximize concurrency, and optimize resource usage. Fluid compute optimizes every request in real-time, routing efficiently, selecting the best compute, and scaling dynamically with a secure infrastructure design that enhances security and enables precise control over workload distribution and scaling decisions. By efficiently managing compute allocation, Fluid compute prevents unnecessary cold starts and scales capacity only when needed, reducing costs by up to 85% for projects enabled with this technology.