Ship multilingual apps 10× faster — without the legacy TMS overhead.
TMS workflows, reimagined for AI coding agents — extract, convert, and sync from your repo. Pay for tokens, not words.
Built for developers who ship · Translators who refine · Localization teams who move fast
Get started for free — no credit card required.
Seamless integration with your favorite i18n library
From
2 weeks
to 1 day
CLI beats
Everything
No mcp, plugins, docs
Intelligent
Prompt suggestions
No docs reading
LLM
Translation
No per-word fees
Pricing
Stored strings = source words × languages. Same base as everyone else. Translation cost = LLM tokens only. That is where we differ.
| Platform | Starter Price | Billing Model | vs i18.dev |
|---|---|---|---|
| Loka•••• | $144/mo | Seat + word-based | 10.3× more expensive |
| Crow••• | $50/mo | Word × language | 3.6× more expensive |
| Local••••• | Custom/mo | Enterprise only | Hidden pricing |
| Tol••• | $53/mo | Per key | 3.8× more expensive |
| Loc••• | $7/mo | Word × language | Word-based |
| SimpleL••••••• | $12/mo | Per key | Fewer strings |
| Phr••• | $27/mo | Per seat | 1.9× more expensive |
| Loca•••• | $34/mo | Per key | 2.4× more expensive |
| i18.dev | $14/mo | Strings + LLM tokens | Best value |
Translation Cost
Everyone charges for stored strings the same way — source words multiplied by languages. But when you actually translate, competitors charge per word. We charge per LLM token.
Crow•••
Per-word MT fees on top of base
Loka••••
150K AI words/yr, overages billed
Tol•••
MT credits per character
Loc•••
Word-based hosting
i18.dev — LLM tokens included
Your plan includes the token budget. Use it however you want — across all languages — without word-count surcharges.
Ran out of monthly tokens? We've got you covered.
Early in a project you often need more tokens than later — bulk extraction, first-pass translations, and multi-language launches add up fast. We can raise your monthly token allowance on any plan so a busy sprint never blocks your ship date.
Developer Experience
Run commands, not docs. Intelligent prompt suggestions watch your project and hand your agent the exact next step — no integration guides, no API rabbit holes.
You confirmed 12 checkout strings in Japanese. Pull them into your repo.
Copy to your agent
Translations are confirmed on my i18.dev dashboard. Sync them into my repo: 1. Run `i18dev pull --confirmed` to download reviewed translations 2. Verify locale files match my next-intl output_dir format 3. Build and spot-check critical flows in each target language
Generated from your recent dashboard activity. No docs required.
# Prompt 1: Set up i18n library $ npx i18dev init --framework next-intl # installs package, creates config # Prompt 2: Convert entire project $ npx i18dev extract --write --translate # scans files, extracts strings, converts to t() # Ship it $ git add . && git commit -m "i18n: conversion" && git push
CLI commands run in milliseconds. MCP has connection overhead, JSON-RPC latency, context switching.
AI agents understand CLI flags, piping, exit codes natively. MCP requires special tooling.
Install once, run anywhere. Cursor, Claude, Codex, Replit — everywhere.
CLI outputs are version-controlled. Your diff is reviewable. MCP calls leave no trail.
Works in Cursor, Windsurf, Zed, Neovim, any terminal. No lock-in.
Commands are precise and deterministic. GUI clicking is slow and error-prone.
Chain in CI/CD, pre-commit hooks, GitHub Actions. Extensions cannot.
Run on servers without displays. The CLI needs nothing but a shell.
Dashboard watches your project — new languages, namespaces, confirmed strings — and generates the exact next prompt for your agent.
Skip integration guides, API references, and setup wikis. Copy one prompt, paste to Cursor or Claude, done.
Added Japanese? Confirmed checkout strings? Prompt suggestions refresh automatically from your recent changes.
Every suggestion includes CLI commands, file paths, and library-specific steps — not generic doc links.
Features
Dashboard tracks your project changes and generates copy-ready prompts for your agent. No documentation reading — just paste and run.
Translations adapt to UI context — checkout CTAs, error messages, onboarding flows — so copy stays natural and accurate in every locale.
Fast, scriptable, git-friendly commands. Your agent runs the CLI — not browser automation, MCP overhead, or extension clicking.
i18dev clien provide tools for your coding agent(Cursor, Claude, or Codex agent.. ) to scans your codebase, finds every hardcoded string, extracts them. No manual grep.
Prompt 1: Set up i18n library. Prompt 2: Convert entire project. Agent handles keys, code replacement, file structure.
No per-word fees. Translate as much as your token budget allows across all languages.
Bring docs, brand materials, and product content — then brainstorm with each translation. The agent researches your target market and competitors to refine tone and messaging.
Up to 20 team members on the free tier. No per-seat pricing.
| Feature | i18.dev | Loka•••• | Crow••• | Tol••• | Loc••• | Simpl•••• |
|---|---|---|---|---|---|---|
| AI coding agent integration | ||||||
| Intelligent prompt suggestions | ||||||
| No documentation reading required | ||||||
| Content-aware localization | ||||||
| Auto string extraction | ||||||
| Auto code conversion | ||||||
| Two-prompt project conversion | ||||||
| CLI-first workflow | ||||||
| Localization agent | Soon | |||||
| Translation dashboard | ||||||
| Translation memory | ||||||
| Glossary management |
Get Started
Free tier: 100 strings, 20 team members, unlimited languages.
Get started for free — no credit card required.