Draft:Lingvanex Translator

Lingvanex is a neural machine translation service launched in March 2017 and owned by Nordicwise Limited, based in Larnaca. Service support translation in 110 languages and can translate text, HTML and files. It offers website interface, API SDK, on-premise solutions and mobile apps for Android, iOS, Windows, MacOS. On-premise versions of Lingvanex can run on Microsoft Windows and Linux. The Lingvanex translation apps are offered free of charge with restrictions that can be removed with a paid subscription.

History
On March 2017 the company Nordicwise Limited was established to offer machine translation services under Lingvanex business name.

On January 2018, Lingvanex launched the first mobile apps for iOS and Android, intended for translation of the text, voice and web pages.

On May 2018 Lingvanex presented their solutions for neural machine translation on mobile phones at European Association for Machine Translation (EAMT).

On April 2020, The translator is recognized by the authoritative publication Entrepreneur as a product, that may be exactly what you need

On April 2020 Lingvanex application for team chat translation was featured by Slack as Brilliant Bot.

On January 2021 Lingvanex application Phone Call Translator was awarded as Product Of Year in Audio and Video category by Product Hunt.

On February 2021, Lingvanex won AWS AI Startup Challenge

On June 2021, Lingvanex integrated machine translation feature into Vivaldi browser

On October 2023, Lingvanex entered the top 10 best translation apps from English to Hindi according to MySmartPrice

On October 2023, Opera for Android integrates Lingvanex translation service

On February 2024, Lingvanex included in the list of the best Google Chrome extension translators according to Android Authority

Technology
For the translator, a sequence-to-sequence model is used, consisting of several layers, each of which includes an encoder and a decoder. Previous versions of Lingvanex used LSTM or GRU (simple recurrent network), attention layers, etc. The current version uses the so-called transformer architecture, when additional self-attention layers are inserted to connect recurrent and convolution networks.

Lingvanex used United Nations and European Parliament documents and transcripts to gather linguistic data and proprietory tool Data Studio to train language models. Most of the translations between language pairs are made via conversion to English as mediary first and then pivot to target language.

Languages

 * 1) Afrikaans
 * 2) Albanian
 * 3) Amharic
 * 4) Arabic
 * 5) Armenian
 * 6) Azerbaijani
 * 7) Basque
 * 8) Belarusian
 * 9) Bengali
 * 10) Bosnian
 * 11) Bulgarian
 * 12) Catalan
 * 13) Cebuano
 * 14) Chichewa
 * 15) Chinese (Simplified)
 * 16) Chinese (Traditional)
 * 17) Corsican
 * 18) Croatian
 * 19) Czech
 * 20) Danish
 * 21) Dutch
 * 22) English
 * 23) Esperanto
 * 24) Estonian
 * 25) Filipino
 * 26) Finnish
 * 27) French
 * 28) Frisian
 * 29) Galician
 * 30) Georgian
 * 31) German
 * 32) Greek
 * 33) Gujarati
 * 34) Haitian Creole
 * 35) Hausa
 * 36) Hawaiian
 * 37) Hebrew
 * 38) Hindi
 * 39) Hmong
 * 40) Hungarian
 * 41) Icelandic
 * 42) Igbo
 * 43) Indonesian
 * 44) Irish
 * 45) Italian
 * 46) Japanese
 * 47) Javanese
 * 48) Kannada
 * 49) Kazakh
 * 50) Khmer
 * 51) Korean
 * 52) Kurdish
 * 53) Kyrgyz
 * 54) Lao
 * 55) Latin
 * 56) Latvian
 * 57) Lithuanian
 * 58) Luxembourgish
 * 59) Macedonian
 * 60) Malagasy
 * 61) Malay
 * 62) Malayalam
 * 63) Maltese
 * 64) Māori
 * 65) Marathi
 * 66) Mongolian
 * 67) Myanmar
 * 68) Nepali
 * 69) Norwegian
 * 70) Odia
 * 71) Pushto
 * 72) Persian
 * 73) Polish
 * 74) Portuguese
 * 75) Punjabi
 * 76) Romanian
 * 77) Russian
 * 78) Samoan
 * 79) Scottish Gaelic
 * 80) Serbian
 * 81) Sesotho
 * 82) Shona
 * 83) Sindhi
 * 84) Sinhala
 * 85) Slovak
 * 86) Slovenian
 * 87) Somali
 * 88) Spanish
 * 89) Sundanese
 * 90) Swahili
 * 91) Swedish
 * 92) Tagalog
 * 93) Tajik
 * 94) Tamil
 * 95) Tatar
 * 96) Telugu
 * 97) Thai
 * 98) Turkish
 * 99) Turkmen
 * 100) Udmurt
 * 101) Ukrainian
 * 102) Urdu
 * 103) Uyghur
 * 104) Uzbek
 * 105) Vietnamese
 * 106) Welsh
 * 107) Xhosa
 * 108) Yiddish
 * 109) Yoruba
 * 110) Zulu