JSON to TypeScript, Zod & Pydantic Converter
Paste any JSON and get TypeScript interfaces, type aliases, a Zod schema, or Pydantic v2 models — nested objects become named sub-interfaces (or nested classes), mixed arrays become unions, and properties missing from some array items are marked optional. Everything runs in your browser.
How it works
The converter parses your JSON with the browser's native parser, then infers a structural type for every value. Nested objects are lifted into named sub-interfaces (named after their property key, PascalCased and singularized for arrays). Arrays of objects are merged across items: a key present in every item is required, a key missing from some items becomes optional (?), and conflicting value types become a union. null values are typed as null and join unions where mixed. Empty arrays are typed unknown[].
Four output styles are supported: interface declarations, type aliases, Zod schemas (z.object constants with an inferred root type), and Pydantic v2 models. Pydantic output uses snake_case field names with Field(alias="...") pointing back to the original JSON key (plus populate_by_name), maps whole-number values to int and decimals to float, and wraps nullable or optional fields in Optional[...]. Your JSON never leaves the page — there is no server and no upload.