AI localization platform for product and engineering teams. Alternative to Lokalise, Phrase and Crowdin. Dashboard, CLI and MCP server.
This repo is the open source (MIT) CLI, SDK and MCP server for i1n.ai. Translation files stay in your repo. Your AI agent extracts strings, pushes them, translates them with your terminology rules and pulls typed keys back. Product, ops and content people edit the copy in the dashboard instead of sending you tickets; the next i1n pull brings their changes into the repo.
Free plan Β· No credit card Β· i1n.ai Β· Docs
- Engineers: keep locale files in git, sync with
i1n push/i1n pull, get typed keys (i1n.d.ts) and a CI check that fails on missing keys or broken placeholders. - AI agents (Claude Code, Cursor, Windsurf, Codex): the MCP server exposes the loop as tools, so "internationalize this component" ends with
t('key')calls and every locale filled in. - Product, ops and content teams: edit copy in every language from the dashboard, with terminology rules, a review queue and roles. A push never silently overwrites their edits (three-way diff).
- Extract: the agent finds hardcoded strings and calls
i1n_extract_and_translate. - Push: keys go to i1n with a three-way diff, so dashboard edits are preserved.
- Translate: AI translation into 182 languages, with placeholder protection, brand voice and terminology rules.
- Pull: locale files plus generated TypeScript types land back in the repo.
- Check:
npx i1n check --min-coverage 95in CI catches drift before it ships.
Works with your current library through Bridge Mode (i18next, next-intl, vue-i18n, react-intl). Formats: JSON, ARB, Android XML, iOS .strings, YAML.
In production at belo, a LatAm fintech with 3.5M users, where around 30 product and ops people use the dashboard.
# To use the CLI (global)
npm install -g i1n
# To use the SDK + types (in your app)
npm install i1n
# Local CLI usage (optional)
npm install -D i1nSupports npm, pnpm, yarn, and bun.
# 1. Initialize (auth + auto-detect setup)
i1n init
# 2. Push your translation keys
i1n push
# 3. Pull translations + auto-generated TypeScript types
i1n pullInteractive setup that prepares your workspace.
- Authenticates via API key.
- New? If you don't have a key yet, the CLI provides clear guidance on how to get started.
- Auto-detects frameworks (Next.js, Vite, Expo, Flutter, Rails, etc.).
- Saves configuration to
i1n.config.json(automatically ignored via.gitignore). - AI Orchestration: Optionally sets up rules for your AI coding tools.
Syncs your local translations to i1n.
- Detects new keys and source changes.
- Smart Translate: Offers to translate missing keys with a cost estimate before proceeding.
- Efficient caching layer β repeated translations cost a fraction of fresh ones.
- Three-way diff β push only sends the (key, lang) pairs you actually changed, never overwriting edits made via the dashboard or by other teammates. See Team workflow for the full conflict model.
Flags:
--translate [langs]β trigger AI translation after push (e.g.--translate es,fr,ja)--strategy <mode>β how to handle real conflicts:interactive(default in TTY),ours,theirs,abort--forceβ shorthand for--strategy ours(overwrite the server with your local values; destructive)
Downloads translations and generates type-safe IDs.
- Updates local locale files in your configured format.
- Generates
i1n.d.tsfor full IDE autocomplete.
Real-time usage tracking.
- View your current plan and credit usage.
- Monitor active language slots and available capacity.
Catch broken translations before they ship. Built for CI.
- Detects missing keys per language, broken interpolation placeholders (
{{count}}lost in translation), empty values, and malformed files. --min-coverage 95fails the build when translation coverage drops below your threshold.--jsonfor tooling. Exit codes:0clean,1errors found,2config problem.- 100% offline β no API calls, no secrets needed in CI.
# .github/workflows/ci.yml
- name: Validate translations
run: npx i1n check --min-coverage 95Turns your IDE into a localization expert.
- Generates project-specific rules for Cursor (
.mdc), Claude Code (CLAUDE.md), Windsurf, and more. - Ensures AI agents follow your naming conventions, file structure, and brand voice.
MCP server for AI coding assistants.
Starts a Model Context Protocol server that lets Cursor, Claude Code, Windsurf, and other AI assistants execute i1n commands directly from your IDE.
# Add to Claude Code
claude mcp add i1n -- npx i1n mcp
# Or add to .mcp.json / cursor config{
"mcpServers": {
"i1n": {
"command": "npx",
"args": ["i1n", "mcp"]
}
}
}9 tools available:
| Tool | Description |
|---|---|
i1n_status |
Get project status, plan, limits, and active languages |
i1n_check |
Validate locale files offline: missing keys, broken placeholders, coverage |
i1n_push |
Push local translation files with three-way diff (preserves server-side edits, aborts on conflict so the agent can resolve) |
i1n_pull |
Pull translations and generate type-safe TypeScript definitions |
i1n_translate |
Translate keys to specified languages using AI |
i1n_add_language |
Add new languages with optional auto-translation |
i1n_extract_and_translate |
Extract strings from code, push as keys, translate to all languages |
i1n_search |
Search existing translation keys by name or value |
i1n_setup_bridge |
Detect your i18n library (i18next, vue-i18n, next-intl, etc.) and wire up i1n bridge mode end-to-end |
Example: tell your AI agent "internationalize this component":
- The agent reads your file and identifies hardcoded strings
- It calls
i1n_extract_and_translatewith the extracted strings - i1n pushes the keys, translates to all active languages, generates types
- The agent rewrites your component with
t('key')calls
i1n is designed for teams where multiple people edit translations in parallel β devs in different branches, copywriters in the dashboard, AI agents via MCP. i1n push is safe to run without worrying that your local working tree might pulverize someone else's edits.
Before every push, the CLI:
- Reads your local locale files (
L). - Asks the server which keys exist and when each was last modified (cheap metadata-only call, ~50Γ smaller than a full pull).
- If anything moved on the server since your last sync, fetches the full server state (
S). - Computes a three-way diff per
(namespace, key, lang)against the last baseline you synced (P, stored inlocales/.i1n-push-state.json).
For each (key, lang) the diff places it in one of these buckets:
| Local | Server | Baseline | Action |
|---|---|---|---|
== server |
β | β | unchanged, skip |
== baseline |
changed | β | server-only β auto-pull into your locale files |
| changed | == baseline |
β | local edit β push it |
| changed | changed | both moved | conflict β resolve interactively |
| missing | present | present in baseline | warn, don't propagate (no delete verb) |
Only the languages that genuinely changed locally are sent. Languages you didn't touch are not in the payload, so the server's per-language merge preserves them. No more "my push silently overwrote yield_rate that I never even opened".
A real conflict means you and someone else both edited the same (key, lang) to different values since the last sync. The CLI shows each one and asks you to pick:
Conflict 1/3: common.greeting [en_us]
βΊ Keep local: "Hello there"
Accept server: "Hi"
Abort push
- Local β push your value, overwrite the server.
- Server β discard your local, auto-pull the server's value into your file.
- Abort β exit; nothing is pushed.
For batch / CI / non-interactive environments, pass a strategy:
i1n push --strategy theirs # accept all server values, push nothing for conflicts
i1n push --strategy ours # local wins (alias: --force)
i1n push --strategy abort # exit on any conflictIn non-TTY contexts (e.g. CI without a strategy flag), push aborts with a diff of the conflicts so you can resolve in code.
If a teammate or someone in the dashboard updated a key you never touched, the server's value is automatically written into your local file at push time and your i1n.d.ts is regenerated if needed. Your working tree ends up reflecting reality β your git diff will show the bring-in so you can commit it alongside your own changes.
The MCP i1n_push tool runs the same diff but defaults to abort on conflict because an AI agent should not silently pick a winner. Conflicts are reported in the response so the agent can decide to pull, ask you, or resolve manually before retrying.
locales/.i1n-push-state.json is gitignored by design β it's working-tree state, like .git/index. On a fresh clone or new branch where the file doesn't exist, the baseline is synthesized from the server. Any local divergence from the server is then treated as a conflict (the CLI can't tell whether you edited locally or have stale data). Run i1n pull first if you just cloned and want to bring everything in cleanly.
| Format | Frameworks | File Sample |
|---|---|---|
| Nested JSON | i18next, next-intl, vue-i18n | en/common.json |
| Flat JSON | React Native, Generic | locales/en.json |
| ARB | Flutter / Dart | app_en.arb |
| YAML | Ruby on Rails | en.yml |
| Android XML | Native Android | strings.xml |
| Apple Strings | iOS / macOS | Localizable.strings |
| TypeScript | Type-safe JSON | locales/en.ts |
The i1n package includes a runtime SDK for web and mobile JS/TS projects. You can use it in two ways:
Use the i1n native engine directly. No external dependencies needed.
import { init, t, setLocale } from "i1n";
// Load your translation resources
init({
locale: "en_us",
resources: {
en_us: {
auth: { login: "Login", title: "Welcome back, {user}" },
items_one: "One item",
items_other: "{count} items",
},
es_es: {
auth: { login: "Entrar", title: "Bienvenido de nuevo, {user}" },
items_one: "Un elemento",
items_other: "{count} elementos",
},
},
});
// Autocomplete and type-safety work out of the box after 'i1n pull'
t("auth.login"); // "Login"
// Support for default values (useful during development)
t("new.key", { defaultValue: "Coming soon..." }); // "Coming soon..."
// Variables & Plurals
t("auth.title", { user: "Fran" }); // "Welcome back, Fran"
t("items", { count: 5 }); // "5 items"
// Switch language at runtime
setLocale("es_es");
t("auth.login"); // "Entrar"Key resolution works with both nested and flat structures automatically β use whatever format your project prefers.
Already using i18next, vue-i18n, or react-intl? Connect it to i1n with one line and get full autocompletion.
import i18next from "i18next";
import { registerI1n, t } from "i1n";
// Set up i18next as usual
await i18next.init({
lng: "en",
resources: {
/* ... */
},
});
// Connect to i1n β one line
registerI1n((key, params) => i18next.t(key, params));
// Now t() uses i18next under the hood, but with strict type checking
t("common.greeting", { name: "World" }); // Powered by i18next, typed by i1nWorks with any library:
- vue-i18n:
registerI1n((key, params) => i18n.global.t(key, params)) - react-intl:
registerI1n((key, params) => intl.formatMessage({ id: key }, params)) - Custom:
registerI1n((key) => myLookup(key))
Define plural variants with _zero, _one, _other suffixes:
// In your translation files:
// "items_zero": "No items"
// "items_one": "One item"
// "items_other": "{count} items"
t("items", { count: 0 }); // "No items"
t("items", { count: 1 }); // "One item"
t("items", { count: 5 }); // "5 items"Three syntaxes supported universally: {var}, {{var}}, %{var}
The SDK works in plain JS β you just don't get autocompletion:
import { init, t } from "i1n";
init({ locale: "en_us", resources: { en_us: { greeting: "Hello {name}" } } });
t("greeting", { name: "World" }); // "Hello World"For a "plug and play" experience, use this minimalist provider pattern.
import { createContext, useContext, useState, useEffect } from "react";
import { init, t, getLocale, setLocale as sdkSetLocale } from "i1n";
// 1. Initialize with wordings
// (In a real app, you'd probably import these from your locales folder)
init({
locale: "en_us",
resources: {
/* ... */
},
});
const STORAGE_KEY = "i1n-locale";
const I1nContext = createContext({
locale: "en_us",
setLocale: (l: string) => {},
});
// 2. Persistent Provider
export function I1nProvider({ children, defaultLocale = "en_us" }) {
const [locale, setLocaleState] = useState(() => {
return localStorage.getItem(STORAGE_KEY) || defaultLocale;
});
// Keep SDK in sync
useEffect(() => {
sdkSetLocale(locale);
}, [locale]);
const setLocale = (newLocale: string) => {
localStorage.setItem(STORAGE_KEY, newLocale);
setLocaleState(newLocale);
};
return (
<I1nContext.Provider value={{ locale, setLocale }}>
{children}
</I1nContext.Provider>
);
}
// 3. Simple Hook
export const useI1n = () => ({ t, ...useContext(I1nContext) });Usage:
const { t, setLocale } = useI1n();
return (
<div>
<h1>{t("auth.title", { user: "Fran" })}</h1>
<button onClick={() => setLocale("es_es")}>EspaΓ±ol</button>
</div>
);Flutter, Android, and iOS projects don't use the SDK. They use the translation files (.arb, .xml, .strings) generated by i1n pull with their native localization systems.
- Auto-Ignore:
i1n initautomatically adds sensitive config files to your.gitignore. - Secret Management: API keys are only stored locally and never committed to version control.
- Encrypted Transmission: All sync operations happen over secure HTTPS channels.
The CLI generates a lightweight declaration file (i1n.d.ts) that automatically augments the i1n package with your project's specific keys.
- Pull: Run
i1n pull. The CLI generateslocales/i1n.d.tsand automatically updates yourtsconfig.jsonso your IDE finds them immediately. - Usage: Import
tfromi1nand get full autocomplete + compile-time checking. No manual path mapping required.
import { t } from "i1n";
// Full autocomplete & compile-time checking
t("auth.login.title");
// ERROR: Argument of type '"auth.login.titlse"' is not assignable...
t("auth.login.titlse");
// Dynamic strings still pass through β useful for `t(item.name)`,
// runtime-built keys, etc.
declare const dynamicKey: string;
t(dynamicKey);Strict literal checking landed in
1.3.0: passing a hard-coded string that isn't inI1nKeysis now a TypeScript error (no more silent[i1n] Missing translationwarnings at runtime). Variables typed asstringkeep working without casts.
| Plan | Price | Keys (shared pool) | Languages | AI translations/mo |
|---|---|---|---|---|
| Starter | $0 | 200 | 2 | 2,000 |
| Pro | $29/mo | 2,000 | 3 | 10,000 |
| Business | $99/mo | 8,000 | 6 | 30,000 |
| Enterprise | From $399/mo | Custom (25k+) | Unlimited | Custom |
CLI, SDK, and MCP server are free on every plan. No credit card required for Starter. Priced per organization, not per processed word. Unlimited viewers on every plan.
Pro lifetime from $199, only for the first 200 users.
MIT β Β© 2026 i1n.ai
