Introduction

translate-kit is an AI-powered translation SDK that generates translations at build time using your own AI models via the Vercel AI SDK. It's compatible with next-intl and supports any AI provider with no vendor lock-in.

Features

• Build-time AI translation — translate JSON message files using any AI model
• Incremental — only new or modified keys get translated via a lock file system
• Any provider — OpenAI, Anthropic, Google, Mistral, Groq, or any Vercel AI SDK provider
• Scanner — extract translatable strings from JSX/TSX source code
• Codegen — replace hardcoded strings with t() calls or <T> components
• Two modes — keys mode (t("key")) or inline mode (<T id="key">text</T>)
• next-intl ready — generates useTranslations-compatible message files

How It Works

translate-kit provides a three-step pipeline:

scan → codegen → translate

1. Scan — parse JSX/TSX files, extract translatable strings, generate semantic keys via AI
2. Codegen — replace hardcoded strings with t("key") calls or <T id="key">text</T> components
3. Translate — diff source messages against a lock file, translate only new/modified keys

You can use any step independently or run the full pipeline. The translate command is the most commonly used on its own — write your source messages manually and let translate-kit handle the rest.

Incremental by Default

A .translate-lock.json file tracks source hashes so re-runs only translate what changed. This keeps API calls fast and costs low.