ModernMT
Machine translation that adapts in real time to your corrections and context.
Standout features
ModernMT’s distinguishing idea is adaptation: it uses the whole document for context and learns in real time from each correction, so it gets better the more you use it.
Worldwide search interest, indexed 0–100 · Google Trends.
ModernMT is the adaptive MT engine built for professionals.
- Developed by Translated, an Italian language-services company in Rome (founded 1999).
- Started as an EU-backed research project; now supports 200 languages.
- Learns in real time from corrections and uses full-document context.
- Recognised by IDC and CSA Research; available via API and in CAT tools like Matecat.
ModernMT’s edge is real-time adaptation.
- Learns from each correction instantly.
- Full-document context per sentence.
- 200-language coverage.
- Integrates with Matecat and Trados.
Trial then API / usage pricing.
ModernMT is for professional translation workflows.
- Professional translators and LSPs.
- Teams with large, ongoing, specialised content.
- CAT-tool users wanting adaptive suggestions.
- Casual users wanting a quick free translator.
- Teams needing a full TMS rather than an engine.
No tool is perfect — the trade-offs to weigh:
- It’s an MT engine, not a full platform.
- Quality skews to European language pairs.
- Best value assumes a professional workflow.
- Adaptation needs corrections to shine.
- ✓Real-time adaptive learning
- ✓Full-document context
- ✓200 languages
- ✓Human-in-the-loop design
- ✓Industry-analyst recognition
- ✕Engine, not a platform
- ✕Skews European
- ✕Best in pro workflows
- ✕Needs corrections to adapt
Professional translators and LSPs value ModernMT for adapting in real time to their corrections and using full-document context, so suggestions improve the more they work. The honest framing is that it’s an MT engine rather than a full platform, and quality skews toward European pairs.
ModernMT is an adaptive neural MT engine from Italian language-services company Translated, supporting 200 languages.
Company figures are drawn from public disclosures and reputable trackers (gathered Jun 2026). User and revenue numbers are estimates and move fast.
Pick up to two other coding tools to see them head-to-head on the same rubric.