Lightpanda ships 1.0 after passing 1.7 million web-platform tests

The machine-focused browser ends its beta period with CORS enabled by default. Its test tally includes a large batch of encoding checks and excludes visual rendering.

By · Published

Primary source: X

Why it matters

Lightpanda's 1.0 release brings its machine-first browser to production, with reported customer use and integrations in agent frameworks. Its 1.7 million-test total reflects broader standards coverage, though many tests come from encoding checks and the browser does not render pages. Adoption depends on whether production browsing jobs need visual output.

A compact server sits beside a monitor displaying a plain browser window with an unmarked webpage.

Lightpanda released version 1.0 on October 2nd, ending the beta period for its browser built for automated web access. The company says the release follows more than 10,000 commits and passes 1,739,845 Web Platform Tests subtests, with cross-origin resource sharing (CORS) enforcement enabled by default.

https://x.com/lightpanda_io/status/2106005006017061163

poster=/api/storage/public-objects/tweet-videos/lightpanda-1-0-production-browser-poster-b5df80a8.jpg|Video from @lightpanda_io on X
Video from the original post on X.

Lightpanda's 1.0 release reflects its view that web automation does not need a browser built to draw pages for people. The open-source browser, written in Zig, runs JavaScript and provides a DOM and web APIs, but omits graphical rendering. That design targets crawlers, search and data extraction systems, tests, and AI agents that need page content and interaction rather than screenshots.

Lightpanda co-founder and CEO Francis Bouvier says the project grew from operating a large-scale web-crawling system at his previous company, where the team collected data from more than 20 million pages a day using headless Chrome. Bouvier previously co-founded BlueBoard, an e-commerce analytics company acquired by ChannelAdvisor in 2020. That experience informed Lightpanda's effort to remove the compute and maintenance overhead of running many Chrome instances for machine-driven browsing.

Infographic showing how Francis Bouvier’s reported experience operating a web-crawling system processing more than 20 million pages a day with headless Chrome informed Lightpanda’s effort to reduce the overhead of many Chrome instances.
Lightpanda says co-founder and CEO Francis Bouvier’s earlier web-crawling experience informed the project’s focus on reducing the overhead of machine-driven browsing — AI explanatory infographic, not documentary evidence. RuntimeWire · AI-generated infographic.

Lightpanda's own account puts the browser at 2,645 subtests when it entered beta in November 2024, then roughly 290,000 by March 2026. A large jump came in April, when the encoding test suite, which contains about 1.15 million subtests, many built from generated character-mapping tables, began passing. The company says it added roughly 340,000 further passes from April through September as it worked on areas including workers, Shadow DOM, IndexedDB, XPath, WebSockets and forms. The tally shows expanding standards coverage, but the raw total gives equal weight to tests that exercise very different parts of a browser.

Lightpanda puts its total at about 80% of Chrome's September 2026 WPT count. The remaining gap is concentrated in CSS, editing and SVG tests tied to layout and painting, according to the release post. Lightpanda does not render pages, take real screenshots or target PDF rendering; teams that need those capabilities still need a full visual browser. Its production claim applies to workloads that can trade visual fidelity for lighter, faster automation, not every task currently handled by Chrome.

The release also makes a security change that matters when a browser executes code from sites an operator does not control: CORS is now on by default. The policy governs whether scripts can read responses from other origins. Lightpanda says browser navigation remains unaffected, while cross-origin fetch requests initiated by page scripts now follow the same restrictions as Chrome. The team recommends keeping the control enabled.

The company points to existing integrations as evidence that users can adopt the engine without replacing their automation stack. Keenable says more than 35% of its AI-search answers depend on JavaScript-rendered pages and reports that Lightpanda makes the work scalable. Vercel's agent-browser can select the engine with a command-line flag, while Nous Research's Hermes Agent supports it as a local browser option. Those examples are company- and partner-reported usage, not independently audited deployment figures.

DeveloperHub.io's founder says moving a prerendering service from headless Chrome to Lightpanda made pages serve four times faster and cut average system load tenfold. The customer-reported result illustrates the operational trade Lightpanda is selling: less infrastructure for workloads that do not need rendered pixels.

Lightpanda's release announcement says some users process more than 30 million pages a day, without identifying them or providing a breakdown. The company also publishes a hosted cloud service alongside its open-source browser. Its current pricing page lists a free plan with 10 browser hours a month and a $19 monthly plan with 300 hours, then $0.08 per additional hour; high-volume and private-cloud terms are custom. Developers can test the hosted service on the free plan, while high-volume and private-cloud terms are negotiated.

The browser is available as version 1.0.0 on GitHub, and Lightpanda says developers can connect it to existing Puppeteer, Playwright, Selenium and ChromeDP workflows. Version 1.0 defines its intended boundary: a production browser for machine-facing jobs where DOM access and JavaScript execution matter more than pixels, with Chrome remaining the fallback for visual work Lightpanda does not do.

Reader comments

Conversation for this story loads after sign-in.