Calque
What it is
Calque is free, MIT-licensed open-source software that turns a single website screenshot into HTML and CSS. It measures pixels to produce either a self-contained HTML file or JSON describing colors, type scale, weights, spacing, container width, gutters, and corner radii. It identifies sections, columns, cards, buttons, alignment, reading order, and vertical gradients, and can crop photographs as assets. Its check command renders the result in an installed Chrome-family browser and compares it pixel by pixel with the screenshot. Calque also ships as a Claude skill that supplies measurements for project code. The workflow uses Tesseract for OCR, with Docker optional. The maker says no model call or API request is made; after OCR data is cached, it works offline. It accepts PNG or JPEG screenshots, requires Node.js 20 or later, and handles one screenshot at a time.
For and against
For
- Free MIT-licensed open source
- Self-contained HTML or JSON output
- Extracts detailed layout measurements
- Pixel-by-pixel browser check
- Offline operation after OCR caching
Against
- Processes one screenshot at a time
- Cannot identify the exact typeface
- Icons become placeholders
- Cropped photos retain screenshot limitations
The verdict
Calque provides a measurement-driven route from one screenshot to editable HTML and CSS. Its offline workflow and validation command are useful, but page coverage and typography remain limited.
Who it suits
It suits developers rebuilding a page from a PNG or JPEG screenshot using plain HTML and CSS custom properties. It also fits users who want local, offline processing after OCR data is cached.
On the record
- Claude integration
- Calque ships as a Claude skill that provides measured screenshot details for Claude to use when writing project code.github.com · read 9 Oct 2026
- Design measurements
- It extracts design tokens including colors, type scale, weights, spacing, container width, gutters, and radii.github.com · read 9 Oct 2026
- Image handling
- With the --assets option, it crops photographs from the screenshot and references those crops in the output.github.com · read 9 Oct 2026
- Inputs and outputs
- It reads PNG or JPEG screenshots and can write a self-contained HTML file or a JSON file of measurements.github.com · read 9 Oct 2026
- Known weak cases
- The maker reports lower reconstruction matches for photograph-heavy pages and identifies dark pages with large glows or gradients behind text as its worst case.github.com · read 4 Oct 2026
- Layout and images
- It finds sections, columns, cards, buttons, alignment, and reading order, and can crop photographs from the screenshot as assets.github.com · read 4 Oct 2026
- Layout features
- It identifies sections, columns, cards, buttons, alignment, and reading order, and can reconstruct vertical gradients.github.com · read 9 Oct 2026
- License
- The repository is MIT licensed, and its README says users may take, host, fork, and sell it.github.com · read 9 Oct 2026
- Limit: icons and photos
- Icons become placeholders, while cropped photographs remain screenshot-resolution images with any overlaid text baked in.github.com · read 9 Oct 2026
- Limit: page coverage
- It reads one screenshot and does not capture scrolling, hover states, or another breakpoint.github.com · read 9 Oct 2026
- Limits
- It processes one screenshot rather than a whole site, cannot identify the exact typeface, and uses cropped screenshot images rather than recovering original photographs.github.com · read 4 Oct 2026
- Limit: typography
- Calque cannot identify a typeface from a screenshot, and font weight resolves to about one step.github.com · read 9 Oct 2026
- Maker
- The repository says Calque is part of Openware, whose GitHub profile describes its work as self-hostable replacements for subscription software.github.com · read 4 Oct 2026
- Other tools
- The documented workflow uses Tesseract for OCR and Docker as an optional way to run Calque in a container.github.com · read 4 Oct 2026
- Output
- It can write a self-contained HTML file or a JSON specification of its measurements.github.com · read 4 Oct 2026
- Performance and accuracy
- The README says processing takes about a second per page and reports fixture match scores ranging from 15.3% to 97.1%.github.com · read 9 Oct 2026
- Privacy and operation
- The maker says Calque makes no model call or API request, and that after its OCR data is cached it works offline; the Docker build preloads that data.github.com · read 4 Oct 2026
- Purpose
- Calque turns a website screenshot into working HTML and CSS by measuring pixels rather than asking a vision model to interpret them.github.com · read 9 Oct 2026
