Skip to content

ai-i18n-toolsTake your single-language project multilingual.

Add languages to your app, website, or repository docs with guided setup and AI translation. Choose your provider and models, and reuse cached translations as your project grows.

ai-i18n-tools logo

Start with the project you have ​

One language today? Start here. That language can be any language: set it as sourceLocale. English is a common example in these guides. Internationalisation (i18n) prepares your project to support different languages; these guides take you from setup to your first translated experience, including the README your repository already shows on GitHub, GitLab, or any other host.

JavaScript or TypeScript app ​

Turn plain-text buttons, labels, and messages into a multilingual UI. Set up your runtime, mark the text to translate, and connect language switching. Includes coding-agent prompts for the preparation step.

React · Next.js · Node.js · CLI

Add i18n to a JS/TS app

Python app ​

Wrap user-facing text and load translated messages with a generated helper. Already on Django or GNU gettext? Keep your gettext calls and let the tool fill the .po files. Run the translation CLI from a standalone executable with no separate Node.js installation.

Streamlit · FastAPI · Flask · Django · CLI

Add i18n to a Python app · Add i18n to a Django / Python gettext app

.NET app ​

Wrap C# and XAML text and load translations from embedded JSON, or keep ResourceManager and write .resx satellites. A dry-run command turns interpolation into T() calls and lists named resource keys, such as Resources.LoginButton, for you to convert.

Avalonia · WPF · MAUI · ASP.NET · Blazor

Add i18n to a .NET app · Add i18n to a .NET ResourceManager app

Website or documentation ​

Publish localised pages and connect them to your site's language routes. Follow framework-specific steps for navigation, theme labels, and the language menu.

VitePress · Starlight · Docusaurus · Astro

Add i18n to a website or docs site

Repository README and docs ​

Translate README.md and other Markdown files into one file per language, with a row of links between them. Pair it with your app's UI strings, or translate the README on its own.

GitHub · GitLab · Bitbucket

Add i18n to your README and Markdown docs

Not sure which path fits? Add i18n to your project lists every walkthrough. See what ai-i18n-tools does for how the pipelines fit together.

From setup to your first new language ​

  1. Prepare your project. Pick your languages, a translation provider, and the models to try. For apps, mark user-facing text and set up the translation runtime. For sites, enable your framework's language routing. The guides include coding-agent prompts where source changes are needed.
  2. Translate into files you can use. Once setup is ready, run sync. It extracts supported app strings, translates the content you enabled, and writes locale files or localised pages.
  3. Load, check, and keep going. Connect the output to your app or site, connect language switching, and check a target language. Rerun after source changes to translate new or changed text.

You handle runtime and routing setup, by hand or with a coding agent. The CLI handles extraction, translation, validation, and file generation—not automatic source rewriting or language-switcher installation. Follow the full quick start.

Your source text. Ready for more languages. ​

This illustrative JS/TS example shows the shape of the result after runtime setup and translation. Marking the source string makes it discoverable; each generated locale file maps that sentence to its translation.

js
t("Save changes");
json
{
  "Save changes": "Änderungen speichern"
}
json
{
  "Save changes": "Salvar alterações"
}
json
{
  "Save changes": "保存更改"
}

The source sentence stays readable in your code. Your app loads the generated locale files at runtime. Want to see the full wiring? Explore the working Next.js example. Websites follow a different path: generate localised pages instead of wrapping every paragraph.

Built for the next update, too ​

Choose your provider and models ​

Use supported providers like OpenAI, Anthropic, OpenRouter, Gemini, local Ollama, or nine other built-in options. You can also use any OpenAI-compatible provider. Translation quality follows the model list (translationModels, plus per-locale overrides). suggest-models, and the same ranking under dashboard Models, builds that list from published translation benchmarks.

LLM providers · AI Models

Reuse work already done ​

Cached translations let reruns focus on new or changed text rather than translating everything again.

Keep it translated

Review AI output locally ​

Inspect translations, correct wording, and manage terminology in the Translation Dashboard.

See the dashboard

Keep app and content together ​

Translate app messages, documentation, nested locale JSON, and SVG text labels from one config. Enable only what you need.

Combining pipelines

See a working project first ​

These examples include configuration and committed translated files. Explore the output without an API key; rerunning translation requires a configured provider and usually a key. Local Ollama needs no API key.

Python in practice ​

See how a Streamlit app wraps its messages and uses translated JSON without adding a JavaScript runtime to the app.

Explore the Streamlit example

Browse all runnable examples, including .NET helpers, plain HTML, other frameworks, and existing-i18n workflows.

Already have i18n? Keep what works. ​

You do not have to start over.

Also have diagrams? Translate SVG text labels alongside your app or docs.

Ready to add your first language? ​

Choose your project for the complete walkthrough, or open the quick start for installation and configuration.

Install the npm package for a Node.js project.

For Python, download the standalone executable—no separate Node.js installation required.

npm version License: MIT

Linux x64 Linux arm64 macOS arm64 macOS x64 Windows x64

Go deeper when you need to ​

Integrations · LLM providers · AI Models · Configuration · CLI reference



About this site

The non-English interface and documentation were translated with ai-i18n-tools. AI translations may contain errors.

All product names, logos and trademarks are the property of their respective owners. Icons and names are used for identification purposes only. This project is not affiliated with or endorsed by those brands.

Released under the MIT License.