Fern vs quicktype

Fern has no write-up on Specifiction yet; its figures are above.

quicktype infers types from structured data and generates models, serializers and validators in more than 25 languages. Inputs include sample JSON, API URLs, JSON Schema, TypeScript and GraphQL queries. It can infer nullability, optional fields, numeric types, dates, UUIDs, enums, unions, maps and nested type names, then generate code to read, write and validate data. The browser app lets users paste JSON and produce code; the project says it works offline and does not send sample data over the Internet. The CLI can process files, URLs, directories of samples or standard input, including in build pipelines, CI jobs and git hooks. A hosted MCP endpoint connects quicktype to Claude Code, Cursor, Codex or other MCP clients without an API key. The project is open source under Apache-2.0, and its web app and CLI are free. The CLI and Node.js packages require Node.js 20 or newer. Platforms listed include web, Linux, macOS, Windows and API. The project directs users to a Slack community for help and discussion. It describes generated types as a way for teams to build API clients, catch API changes and give IDEs and AI agents a typed model of data.