Next.js 16 Demo's Speed Comes at a Cost; Vercel Announces Official Bun Runtime Support

A recently unveiled demo website, showcasing features of an upcoming ‘Next.js 16 Faster’ version and promoted for ‘exceptional performance,’ has sparked significant debate within the developer community. While initial user experience highlights ‘incredible speed’ on navigation, scrutiny reveals the site employs highly aggressive prefetching and preloading techniques. Critics report downloads of up to 66 megabytes of data upon initial interaction and subsequent scrolling, attributing the perceived speed to extensive background data loading rather than inherent framework optimizations. Concerns are particularly high for mobile users, where such data consumption could severely impact data plans. Furthermore, some argue that this prefetching strategy, a common technique across frameworks (e.g., Astro also supports prefetch options), is not a unique benefit of Next.js and could disproportionately increase bandwidth costs for Vercel, the hosting provider, as data is served even if not viewed by the end-user. The consensus among critics suggests removing such extreme preloading to accurately demonstrate Next.js’s true performance capabilities.

In related news, Vercel has officially announced robust support for the Bun 1.x runtime environment. This marks a significant milestone, as previously, while bun install was feasible, deploying projects leveraging Bun as the primary runtime on Vercel was largely unsupported or required unofficial workarounds. Developers can now enable Bun support by simply adding "bunVersion": "1.x" to their vercel.json configuration. This integration is expected to yield substantial benefits, including reduced memory consumption, faster response times, and consequently, lower computation costs for projects hosted on Vercel. Beyond performance enhancements, official Bun support also enables seamless access to Bun’s native APIs, such as S3 and SQL clients, further extending the capabilities for Vercel-hosted applications. While initial tests by some users on complex projects encountered deployment errors, the official support is anticipated to stabilize and provide a powerful new option for optimizing Vercel deployments.