Streamdown 1.6 is now available to run faster and ship less code
Vercel ships Streamdown 1.6 with 30% faster performance and smaller bundles — a move that matters for AI-powered content platforms handling real-time markdown rendering.
Why it matters
Streamdown 1.6 optimizes the infrastructure layer for rendering AI-generated content at scale. For founders building AI products that stream markdown output (reasoning chains, code generation, documentation), this reduces latency and bandwidth costs — directly improving the UX of agentic workflows.
The key facts
12 to knowPerformance improvements via memoization, LRU caching, optimized string operations
Bundle size reduction through regex removal and custom markdown renderer
Lazy-loading for Code Blocks, Mermaid, and Math components
Static Mode for non-streaming markdown rendering to reduce overhead
Code highlighting system rebuilt with simpler tokenization and line numbers
Diagram export as SVG, PNG, or source code with fullscreen zoom/pan
Streamdown 1.6 reduces bundle size through memoization, LRU caching, optimized string operations, and regex removal
Lazy-loading for Code Blocks, Mermaid, and Math components via React.lazy() and Suspense
Custom markdown renderer replaces React Markdown, reducing streaming overhead
Static Mode added for rendering markdown without streaming
Code highlighting system rebuilt with new tokenization approach including line numbers
Diagrams export as SVG, PNG, or source code with fullscreen zoom/pan controls
Go to the source
Vercel Blogvercel.com
Publisher excerpt: Vercel Streamdown 1.6 is now available with major improvements to performance, bundle size, and the authoring experience. Streamdown now runs faster and ships less code thanks to memoization, LRU caching, optimized string operations, and the removal of regexes. Several product enhancements include:…