What is lingo?

Lingo is a library that turns messy, natural language text into clean, standardized values that a database or program can trust. People type things like 180cm, 5ft 11, or next friday into form fields, and AI models often output oddly formatted strings like 5'11" or twenty-five kg. Lingo sits between that messy input and your system, parsing quantities, units, dates, and ranges into one consistent number and unit, then can turn that clean value back into readable text again. On the web side, it lets a single text field replace a separate value box and unit dropdown, accepting however people actually type measurements, including typos, and offering did you mean suggestions rather than harsh error messages while someone is still typing. When input is genuinely ambiguous, like whether 1,234 means one point two three four or one thousand two hundred thirty four, it returns its best guess along with alternatives and a warning instead of silently picking one. It ships as a headless controller, a custom input element, and a React hook, with no visual styling forced on you. For AI tools, lingo provides ready made field definitions that plug into popular AI development kits and protocols, so a language model's tool call gets validated and corrected using the same underlying parser used for human typed forms. It rejects risky ambiguous values at that boundary with clear error messages a model can use to fix its own mistake on the next attempt, and results are reproducible when you supply a fixed reference time, which matters for retried or queued requests. The library itself has zero runtime dependencies and includes English support by default, with optional language packs for several other languages. This repository is a monorepo containing the published library package, along with a documentation site and internal planning files. This project is aimed at developers building web forms or AI powered tools who need reliable, consistent parsing of everyday quantities and dates. It is released under the MIT License.