Ladybird Browser Adds Inline PDFs, Faster HTML Parsing

Key Takeaways

• Ladybird now renders PDFs inline using pdf.js with full navigation, zoom, and text selection
• Speculative HTML parsing fetches resources before the main parser reaches them, cutting load times
• Off-thread JavaScript compilation moves roughly 200ms of work off the main thread on YouTube alone

April by the Numbers

Ladybird merged 333 pull requests in April 2026 from 35 contributors. Seven of those contributors made their first commit to the project. The independent browser, which accepts no advertising or corporate investor money, also picked up two new sponsors: the Human Rights Foundation contributed $50,000 through its AI for Individual Rights program, and individual sponsor Jakub Stęplowski added $1,000.

The technical updates focus on three areas: document rendering, parsing speed, and JavaScript execution. Each targets a real bottleneck in modern web browsing.

Inline PDF Viewer via pdf.js

PDFs now render directly in the browser window instead of prompting a download or external app. Ladybird bundles Mozilla's pdf.js, the same viewer Firefox uses. It handles page navigation, text selection, zoom, and find-in-document search. The implementation is entirely JavaScript, HTML, and CSS.

The team used Intel's ISA Manual as a stress test. Profiling pdf.js loading that document revealed opportunities to improve Ladybird's typed-array view cache and :has() selector invalidation. Both Qt and AppKit user interfaces now support the feature.

Browsing History Gets Smarter

The address bar now offers history-aware autocomplete. Type a few characters and you get suggestions showing previously visited pages with their favicons and titles, plus search engine shortcuts and plain URL completions.

Under the hood, a SQLite-backed HistoryStore saves every navigation with its title, favicon, visit count, and timestamp. The Privacy settings page includes a working "Clear browsing history" button. Both the Qt and macOS AppKit interfaces render the new rich suggestion rows.

Speculative and Incremental HTML Parsing

Two parsing changes should make pages feel faster. First, the HTML parser now processes the response body as it arrives rather than waiting for the complete download. Bytes stream through a text decoder into the tokenizer chunk by chunk. The tokenizer pauses when it runs out of input and resumes when more shows up.

Second, Ladybird now has a speculative HTML parser. When the main parser blocks on a synchronous external script, a separate tokenizer scans ahead through unparsed input. It issues speculative fetch requests for resources it finds: script src, stylesheet links, preloads, and images. It correctly handles base href tags, template content, and foreign content.

A follow-up patch connected the speculative parser to the document's preload map. Resources discovered speculatively get deduplicated against later fetches from the main parser. This prevents double-requesting the same file.

Off-Thread JavaScript Compilation

Bytecode generation for fetched scripts now runs on a background thread pool. Worker threads produce the bytecode and the data needed to build an Executable object. Everything that touches the VM or garbage collector heap stays on the main thread.

This covers classic scripts, ES modules, and top-level immediately-invoked function expressions. On YouTube alone, the change shifts roughly 200ms of work from the main thread to background threads. That's time the main thread can spend rendering or responding to user input.

Per-Navigable Rasterization

Each Navigable, which represents a browsing context like a tab or iframe, now rasterizes independently on its own thread. This means one slow-rendering iframe won't block the main document's paint operations. The change should improve perceived performance on complex pages with multiple embedded contexts.

Why This Matters for Browser Competition

Ladybird is building a browser engine from scratch without using Chromium's Blink, Firefox's Gecko, or WebKit. That's a multi-year undertaking. Most of the work involves implementing web standards piece by piece and optimizing the resulting code.

The April updates show the project moving past basic functionality toward performance work. Speculative parsing, off-thread compilation, and parallel rasterization are the kinds of optimizations that separate usable browsers from fast ones. Chrome and Safari have had these features for years. Ladybird is catching up.

The sponsorship model is also worth watching. The Human Rights Foundation's contribution through an AI rights program suggests some funders see independent browsers as infrastructure for digital rights. No Google analytics baked in. No ad-tech partnerships. Just a browser that loads web pages.

Frequently Asked Questions

What is Ladybird browser?

Ladybird is an independent, open-source web browser being built from scratch without using existing browser engines like Chromium or Gecko. It's funded entirely by donations and sponsorships.

Can I use Ladybird as my main browser today?

Not yet. Ladybird is still in active development and lacks many features and optimizations needed for daily use. The project publishes monthly updates tracking its progress toward a usable release.

What does speculative HTML parsing do?

When the main parser blocks on a script, a separate tokenizer scans ahead to find resources like stylesheets and images. It requests them early so they're ready when the main parser catches up, reducing page load time.

How is Ladybird funded?

Ladybird accepts sponsorships from companies and individuals. In April 2026, the Human Rights Foundation contributed $50,000 and individual sponsors added $1,000. The project does not accept advertising or venture capital.

Source: Hacker News: Best