Lara Translate, the AI-powered translation platform for developers and enterprises, today announced its selection by Scale AI as one of just 36 real-world Model Context Protocol (MCP) servers in the MCP Atlas benchmark. Accounting for 7% of all evaluation tasks, Lara is now established as the de facto translation layer for large language model (LLM) workflows.
Quick Intel
Translation Infrastructure for the AI Agent Era
The inclusion in Scale AI's benchmark reflects Lara's role as a foundational building block in agentic AI systems. To complement this recognition, Lara shipped an MCP Glossary Management Toolkit in March, enabling developers to manage terminology assets directly inside AI workflows. This makes Lara the only translation MCP server with native, in-workflow glossary control.
"When Scale AI builds a benchmark to measure how well AI agents use tools, and translation is 7% of it, that tells you everything about the industry's direction, Lara is already there." — Maurizio Tiberi, Senior Director, Lara Translate
A Complete Multimodal Translation Platform
March also marks the completion of Lara's full multimodal SDK rollout, offering text, voice, and image translation through a unified API:
Expanding the Product Ecosystem
Beyond the API, March updates included table translation directly in the UI and a major rewrite of the Lara Browser Extension for complex web structures across Gmail and Google Docs. Additionally, the Google Sheets add-on now supports SSO, while mobile apps gained native OS-level document sharing.
About Lara Translate
Built on 20 years of research at Translated, Lara Translate delivers context-aware, memory-augmented translations via a developer-first API and suite of productivity tools.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones—and thrive on the tech pulse.