Lara Translate

You can connect AI applications to Lara Translate's translation capabilities using the Lara Translate MCP Server. This open protocol-based server enables language detection, context-aware translations, and translation memories to be accessed by external AI tools, agents, or workflows in a standardized way.

How to use

You use the Lara Translate MCP Server by configuring your MCP-enabled client to connect via one of the supported protocols. You can run the server remotely over HTTP or locally via STDIO. Once connected, your AI application can discover the available translation tools, submit translation requests, and receive translated results for integration into its responses.

How to install

Prerequisites: you need a supported runtime for the chosen connection method. For HTTP access, you only need network access to the remote MCP URL. For STDIO usage, you need Node.js installed or Docker installed to load and run the MCP locally.

HTTP remote server setup (recommended): use the provided remote endpoint and credentials in your MCP client configuration.

STDIO local setup (NPX): ensure Node.js is installed, then configure your MCP client to load the MCP via NPX as shown in the snippet.

Security notes

When running your own HTTP server instance (not using the remote https://mcp.laratranslate.com/v1), all connected clients share the same Lara API credentials configured via LARA_ACCESS_KEY_ID and LARA_ACCESS_KEY_SECRET. For multi-tenant scenarios, use the remote server where each client provides credentials via headers, or deploy separate MCP server instances per user with isolated credentials.

Verify installation

After restarting your MCP client, confirm Lara Translate MCP appears in the list of available MCPs. Then translate a simple phrase to ensure the flow works end-to-end, and observe the client’s response indicating Lara Translate is in use.

Popular clients that support MCPs

The MCP ecosystem supports several clients. You can connect the Lara Translate MCP Server to clients that support URL-based MCP configuration or direct command-based configuration.

Notes on usage with different protocols

HTTP Server usage enables remote connections via URL and headers. STDIO Server usage enables local, self-contained operation through runtime commands such as NPX or Docker. Each method includes its own credential handling as shown in the configuration examples.

Available tools

translate

Translate text blocks between languages with optional context, glossaries, and privacy controls; returns translated blocks preserving the original structure.

list_glossaries

List all glossaries available for enforcing terminology in translations.

get_glossary

Retrieve a specific glossary by its ID to inspect its terms and usage.

list_memories

List saved translation memories for reuse across translations.

create_memory

Create a new translation memory, optionally importing from an external source.

update_memory

Rename or rename/update a translation memory.

delete_memory

Delete a translation memory by its ID.

add_translation

Add a translation unit to one or more memories, including source, target, and context sentences.

delete_translation

Remove a translation unit from a memory by identifying its source, target, and context.

import_tmx

Import a TMX file into a specified translation memory, optionally gzip-compressed.

check_import_status

Check the status of an ongoing TMX import operation.