Translated

What Is Translated?

In 1999, when machine translation meant stilted, often comical word-for-word conversions, Marco Trombetti and Isabelle Andrieu founded Translated in Rome with a conviction that would take decades to prove correct: the future of translation would be human and machine working together, not one replacing the other.

Twenty-five years later, that conviction has materialised into one of Europe's most significant language technology companies. Translated combines its proprietary AI translation engine — ModernMT — with a network of over 200,000 professional translators to deliver a hybrid workflow that pure machine translation cannot match in quality, and pure human translation cannot match in speed. The result is enterprise-grade localisation that serves organisations including Google, Airbnb, and the European Commission.

What makes Translated distinctive is not just the hybrid approach — it is the feedback loop. ModernMT is an adaptive neural machine translation engine that learns from translator corrections in real time. Every human edit improves the machine's output for the next translation. This creates a compounding quality curve: the more you use the system, the better it gets at your specific content and terminology. Over 25 years, across 200+ language pairs, this feedback loop has accumulated an extraordinary depth of linguistic data.

Translated also created Matecat, an open-source computer-assisted translation (CAT) tool used by translators worldwide, and founded Pi School, an AI academy in Rome. The company operates from Italy, processes data within the EU, and maintains GDPR compliance across all services.

Key Features

ModernMT Adaptive Translation Engine

ModernMT is Translated's core technology innovation. Unlike static machine translation engines that produce the same output regardless of context, ModernMT adapts in real time to your content, terminology, and style preferences. When a translator corrects a machine suggestion, ModernMT incorporates that correction immediately — improving subsequent translations in the same session and building a persistent memory of your linguistic preferences.

The engine supports 200+ language pairs and can be accessed via API for self-service machine translation, or as part of the full human-in-the-loop workflow. For organisations with specific terminology (legal, medical, technical), the adaptive capability means the engine becomes increasingly accurate over time without manual glossary maintenance.

Hybrid AI + Human Workflow

Translated's standard professional translation workflow follows a defined process: ModernMT produces the initial translation, a matched professional translator post-edits it, and quality assurance checks are applied. The result is translation quality that exceeds pure machine output while being delivered faster and more cost-effectively than fully manual translation.

T-Rank, Translated's AI-powered translator matching system, selects the optimal translator for each project based on language pair, subject matter, past performance, and availability. This is not random assignment — it is algorithmic optimisation that ensures the right human expert reviews the right content.

Matecat Open-Source CAT Tool

Translated created and maintains Matecat, a free, open-source computer-assisted translation tool that integrates with ModernMT. Matecat provides translation memory management, terminology management, quality assurance checks, and real-time collaboration features. Its open-source nature demonstrates Translated's commitment to the broader translation community and provides transparency into how translations are processed.

Enterprise Localisation Platform

For large-scale localisation projects, Translated offers a project management dashboard, real-time progress tracking, file format support for over 50 formats, and dedicated project managers. Volume discounts, custom SLAs, and dedicated workflows accommodate the complexity of enterprise localisation — multi-market launches, continuous content updates, and regulated content requirements.

ModernMT API

For self-service use cases where speed matters more than perfect quality — internal communications, content discovery, user-generated content moderation — the ModernMT API provides direct access to the machine translation engine without human review. The API supports adaptive learning, custom model training, and is available as both a cloud service and an on-premise deployment for organisations with strict data sovereignty requirements.

Pricing

Translated's pricing reflects its position as a professional services provider rather than a self-service SaaS tool. The ModernMT API operates on a pay-per-character model for self-service machine translation. Professional translation with human post-editing is priced per word, varying by language pair and turnaround time. Enterprise pricing is custom and includes volume discounts, dedicated project managers, and SLA guarantees.

The lack of transparent, self-service pricing is the most common criticism of Translated's commercial model. Most projects require a custom quote, which adds friction to the evaluation process. For organisations accustomed to the instant, self-service pricing of DeepL or Google Translate, this enterprise-oriented approach can feel slow.

The cost per word for human-reviewed translation is naturally higher than pure machine translation services. The value proposition is quality — for business-critical content (contracts, marketing copy, regulatory documents), the accuracy improvement justifies the premium. For high-volume, lower-stakes content, the ModernMT API alone may be sufficient.

EU Compliance & Privacy

Translated S.r.l. is headquartered in Rome, Italy, and processes all data within the EU. The company maintains GDPR compliance, offers NDA and confidentiality agreements for sensitive content, and holds quality management certifications.

For organisations translating sensitive documents — legal contracts, healthcare records, financial reports — Translated's EU data processing and confidentiality framework provides the assurances needed. The on-premise ModernMT deployment option adds another layer for organisations with the strictest data sovereignty requirements.

The 200,000+ translator network introduces a distributed data access model, which Translated manages through contractor agreements and access controls. Organisations with particularly sensitive content should discuss the specific data handling procedures for their project type.

With an EU compliance rating of 8.5, Translated scores well for the translation services category.

Who It's Best For

Enterprises localising content at scale across multiple markets and languages, who need quality that exceeds pure machine translation but with faster turnaround than fully manual processes.

Legal, medical, and financial organisations translating regulated content where accuracy is non-negotiable and EU data processing is mandatory.

Product teams needing continuous localisation integrated into their development pipeline via the ModernMT API, with human review available for customer-facing content.

Organisations with specialised terminology that benefit from ModernMT's adaptive learning — the engine improves with each project, reducing cost and improving quality over time.

The Verdict

Translated represents what happens when you invest 25 years in the intersection of human expertise and machine intelligence. The ModernMT adaptive engine is genuinely innovative, the hybrid workflow delivers quality that pure automation cannot match, and the EU data processing and open-source contributions (Matecat) demonstrate institutional integrity. The pricing opacity and longer turnaround times compared to instant machine translation are real trade-offs. But for organisations where translation quality directly impacts business outcomes — and for whom "close enough" is not acceptable — Translated is one of Europe's most credible language technology partners.

Frequently Asked Questions

How does Translated differ from DeepL or Google Translate?

Translated combines AI machine translation with professional human post-editing, delivering higher quality than pure machine translation for business-critical content. DeepL and Google Translate are fully automated services with no human review step. Translated is the choice when accuracy matters more than instant delivery.

What is ModernMT?

ModernMT is Translated's adaptive neural machine translation engine. It learns from translator corrections in real time, continuously improving translation quality for your specific content, terminology, and style. It supports 200+ language pairs and is available via API or as part of the full translation workflow.

Does Translated offer pure machine translation without human review?

Yes. The ModernMT API provides direct access to the machine translation engine for use cases where speed matters more than perfect quality — internal communications, content discovery, or prototyping. Human post-editing can be added for customer-facing or regulated content.

How many languages does Translated support?

Translated supports 200+ language pairs through its network of 200,000+ professional translators, covering virtually all commercially significant language combinations.

Is Translated GDPR compliant?

Yes. Translated S.r.l. is headquartered in Rome, Italy, processes data within the EU, and complies with GDPR requirements. NDA and confidentiality agreements are available for sensitive content. An on-premise deployment option is available for the ModernMT engine.