MCP Local Translate runs on your own machine — the client starts it, so there is no endpoint to ping. 60 installs a week from npm. Last commit 12 Jul 2026.
Local NLLB translation for 200+ languages. Zero cloud, zero cost, runs entirely on your machine.
Today is the operative word: we check MCP Local Translate every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add mcp-local-translate -- npx -y @damoqiongqiu/mcp-local-translate
{
"mcpServers": {
"mcp-local-translate": {
"args": [
"-y",
"@damoqiongqiu/mcp-local-translate"
],
"command": "npx"
}
}
}
[mcp_servers.mcp-local-translate]
command = "npx"
args = ["-y", "@damoqiongqiu/mcp-local-translate"]
{
"mcpServers": {
"mcp-local-translate": {
"args": [
"-y",
"@damoqiongqiu/mcp-local-translate"
],
"command": "npx"
}
}
}
{
"mcpServers": {
"mcp-local-translate": {
"args": [
"-y",
"@damoqiongqiu/mcp-local-translate"
],
"command": "npx"
}
}
}
This one needs environment variables set before it will start:
HF_ENDPOINT (HuggingFace endpoint override for downloading models (e.g. https://hf-mirror.com for China).), HF_AUTO_MIRROR (Auto-detect HuggingFace mirror for China. Set to '0' or 'false' to disable.), CACHE_DIR (Directory where Transformers.js models are cached (defaults to ./models/).), TRANSLATE_DEVICE (Execution device: 'cpu' (default) or 'webgpu' for GPU acceleration.), HTTPS_PROXY (Proxy URL for model downloads. Also supports HTTP_PROXY and lowercase variants.).
The author declared them in the registry entry; get the values from the project itself.
Search local email archives (.mbox/.eml) entirely on your machine. No OAuth, no cloud.
Localization for AI agents: projects, languages, glossaries and translations from your agent
Local memory for AI agents. One SQLite file on your machine — no cloud, no account.
MacWhisper transcription for Claude Desktop. Fully local, no cloud, audio stays on your Mac.
Offload cheap work from your AI agent to a local Ollama model, at zero cloud cost.
Local secret & credential checks for AI-assisted dev — runs on your machine, results redacted.
Zero-egress PII redaction for Claude. Runs locally — no data leaves your infrastructure.
Local-first long-term memory for AI coding agents. Stored on your own machine.
Answers built from our own checks of this server.