AI Interpreting as a service

We, the Charles University team of ELITR, are ready to serve private or public events with “AI Interpreting” — live translation from 56 spoken languages with automatic language identification into 43 target languages of text. We can serve your event of any kind, an online or face-to-face meeting with sound recording.

ELITR AI Interpreting can help you easily expand your target audience group, and much more. For example, you can invite your foreign partners to your event and help them understand via AI Interpreting.

How can you use our service?

  1. Contact us and let us know:
    • When is your event and how long?
    • The source and target languages you need.
    • How can we connect the speech at your event? Either through an online meeting app (creates a delay but is easy to set up), or through an operator on your side (faster, but must be instructed).
    • Do you need a guarantee of confidentiality? Options: No/password protected/NDA.
  2. (Optional) Test the service before your event.
  3. Use the service at your event.
    • Deliver your high-quality audio to our premises.
    • We process it and publish translations on our web page.
    • Your event participants read translations on the web, either on their own devices or on your big shared screen.
  4. Please, give us feedback.
  5. (Optional) Remuneration if you liked it.

Demo video

This video shows a translation from Czech into English and German by ELITR in 2020. Now, in 2024, we use the same output web interface, but we have much better translation models for more languages.

You can read and see other ELITR demos here, here, here, and in Czech Radio here.

Supported languages

Our service supports the following 56 source speech languages:

Afrikaans, Arabic, Armenian, Azerbaijani, Belarusian, Bosnian, Bulgarian, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, Galician, German, Greek, Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Italian, Japanese, Kannada, Kazakh, Korean, Latvian, Lithuanian, Macedonian, Malay, Marathi, Maori, Nepali, Norwegian, Persian, Polish, Portuguese, Romanian, Russian, Serbian, Slovak, Slovenian, Spanish, Swahili, Swedish, Tagalog, Tamil, Thai, Turkish, Ukrainian, Urdu, Vietnamese, and Welsh.

We are ready to deliver translations into the following 43 target languages in parallel. We support all 24 EU languages and 19 more.

  • The EU languages are:
    Bulgarian, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Irish, Italian, Latvian, Lithuanian, Maltese, Polish, Portuguese, Romanian, Slovak, Slovene, Spanish, and Swedish.
  • Other supported languages are:
    Albanian, Arabic, Armenian, Azerbaijani, Belorussian, Bosnian, Georgian, Hebrew, Icelandic, Kazakh, Luxembourgish, Macedonian, Moldovan, Montenegrin, Norwegian, Russian, Serbian, Turkish, and Ukrainian.

If you need another language pair, let us know. We may add it.

Background

ELITR — European Live Translator, was an EU-funded project for live speech translation research and innovation in which a platform for speech translation of live events was developed. Currently, in 2024, the Charles University team uses ELITR platform with Whisper model for transcription and translation into English. We adapted it for real-time mode with Whisper-Streaming.

Team: Dominik Macháček, Ondřej Bojar, Dávid Javorský, Sangeet Sagar.

We are affiliated with ÚFAL — Institute of Formal and Applied Linguistics, Faculty of Mathematics and Physics, Charles University.

Address: Charles University, Faculty of Mathematics and Physics, Malostranské náměstí 25, 118 00 Praha 1, Czech Republic. VAT ID: CZ00216208.

Contact

Please contact Dominik Macháček by email: machacek@ufal.mff.cuni.cz .