About us
Creation Story
The project was created so that people with hearing impairments can watch and understand video content. According to WHO data, approximately 430 million people experience hearing loss, which accounts for roughly 6% of the global population. In Russia, 13 million people live with hearing impairments (1 million of whom are under 18 years old). In Latvia, there are 32 thousand such people, in Estonia – 28 thousand, and in Lithuania – 19 thousand.
Initially, the portals Movify.TV and Kids.Movify.TV (for children) were created, where people with hearing impairments could watch movies and cartoons for FREE. A special (new) type of subtitle was developed, which, in addition to actors’ lines, also indicates the gender of the speaker and their role in the film. Initially, these subtitles were created manually. Subsequently, the development of a universal AI platform began for the automatic conversion of any video into subtitles for people with hearing impairments. This platform was named MultiVoice.AI
During the platform’s development, a television company from Latvia approached us and proposed creating 24/7 subtitles in Latvian for three TV channels, as well as AI voiceovers for films in Latvian, Estonian, and Lithuanian. Later, a Russian television company came to us with a similar request. In 2025, MultiVoice.AI is creating nearly 1000 hours of subtitles and 250 hours of voice-dubbed video (films) monthly. Furthermore, MultiVoice.AI is developing its own speech synthesis server for various languages.
Legal Information
Movify.CY Ltd. is registered in the Republic of Cyprus Registrar of Companies on June 13, 2024, with registration number ΗΕ461401 and taxpayer identification number 60077645I. The company’s registered and postal address is Agia Zoni, 30a, Fantaros Court, Flat/Office 303, 3027, Limassol, Cyprus. The company is a tax resident of the Republic of Cyprus. The company holds specialized innovative IP Box status in the Republic of Cyprus. Company details can be verified free of charge on the Republic of Cyprus Registrar of Companies website here.
Our team
Our team is focused on developing AI for 100% automation of the localization process (including for people with hearing impairments) of any video or audio content (from any language to any other language). We are developing AI that can translate any content with 100% accuracy. Currently, 12 people from various countries are involved in the project’s creation:
1. Team Lead with 20 years of experience in creating and managing media projects (Cyprus).
2. Client Application Programmer for Dubbing Studio (Spain).
3. Speech Synthesis and Voice Cloning Server Programmer (Ukraine).
4. Diarization Server Programmer (Spain).
5. Sound Engineer (Latvia).
6. Philologist (Latvia).
7. 4 Data Annotators for AI training (Latvia, Russia).
8. Network Administrator (Poland).
9. Lawyer with experience in intellectual property rights (Russia).
What We Do
The current development of AI does not yet allow for fully entrusting artificial intelligence with the localization (translation) of content (video) from one language to another
Artificial intelligence cannot:
- Recognize and understand with 100% accuracy what is being said in a video. Noise, music, and special effects prevent even the best speech recognition systems from 100% understanding what is being said in a film. Due to this, errors and ‘hallucinations’ appear in speech recognition (transcription) systems.
- No speech recognition system can 100% accurately identify actors’ voices. Furthermore, in films, two or more characters may speak simultaneously, which further complicates recognition.
- Artificial intelligence cannot correctly (or naturally) translate idioms, phrasal verbs, and jokes.
- An automatic video translator speeds up or slows down speech to fit the translated phrase into the allotted time. Because of this, the translated video sounds unnatural.
- For languages spoken by a small number of people (up to 10 million, such as Latvian, Lithuanian, and Estonian), there are currently no speech synthesis models that speak naturally, conveying human emotions. While many speech synthesis systems exist for major languages (spoken by hundreds of millions of people – for example, Chinese or English), such systems are absent for ‘smaller’ languages
- Furthermore, processing video requires expensive and energy-intensive Nvidia servers. The use of energy-intensive Nvidia servers significantly pollutes the environment.
Our team is solving all these problems. We are developing AI models for smaller languages so that residents of countries with ‘smaller’ populations can benefit from AI technologies. This includes enabling people with hearing impairments to watch and understand videos just like everyone else. We use Nvidia GPUs with minimal electricity consumption to reduce environmental pollution.
Comparison of Microsoft Azure and MultiVoice.AI
Microsoft Azure offers the use of only 2 voices for Latvian (lv-LV-EveritaNeural, lv-LV-NilsNeural), only 2 voices for Estonian (et-EE-AnuNeural, et-EE-KertNeural), only 2 voices for Lithuanian (lt-LT-OnaNeural, lt-LT-LeonasNeural), only 3 voices for Russian (ru-RU-SvetlanaNeural, ru-RU-DmitryNeural, ru-RU-DariyaNeural), and only 2 voices for Ukrainian (uk-UA-PolinaNeural, uk-UA-OstapNeural). Microsoft Azure does not have programs for automatic video voiceover/dubbing.
MultiVoice.AI offers 24 AI voices for Latvian, 24 AI voices for Russian, 24 AI voices for Lithuanian, 24 AI voices for Estonian, and 24 voices for Ukrainian. MultiVoice.AI has developed a program for AI video voiceover/dubbing.
Comparison of Google and MultiVoice.AI
Google offers the use of only 1 voice for Latvian (lv-LV-Standard-B) and only 1 voice for Lithuanian (lt-LT-Standard-B). Estonian language is not available. For Russian, there are 18 voices, and for Ukrainian, there are 32 voices.
MultiVoice.AI offers 24 AI voices for Latvian, 24 AI voices for Russian, 24 AI voices for Lithuanian, 24 AI voices for Estonian, and 24 voices for Ukrainian.
Comparison of Elevenlabs and MultiVoice.AI
Elevenlabs offers the use of only one voice for Latvian (jessica), only one voice for Lithuanian (jessica), and only one voice for Estonian (jessica). For Russian, there are 5 voices, and for Ukrainian, there are 4 voices.
MultiVoice.AI offers 24 AI voices for Latvian, 24 AI voices for Russian, 24 AI voices for Lithuanian, 24 AI voices for Estonian, and 24 voices for Ukrainian.
Our Competitors
Rask.ai
Rask.AI Features: Automatic video and audio translation with API scalability. According to the Rask.AI website, 100 minutes of video translation to another language costs 120 USD. For Latvian, Lithuanian, and Estonian languages, there are unemotional AI voices from Microsoft, with only 2 voices available for each language.
Comparison with MultiVoice.AI: MultiVoice.AI offers 24 voices for Latvian, Lithuanian, and Estonian. The cost for translating 100 minutes of a film to another language with MultiVoice.AI is 80 Euros (or 93.78 USD). With MultiVoice.AI, users can also edit the quality of the translation and voiceover themselves, similar to Rask.AI. MultiVoice.AI has an additional option: human verification of translation quality by a professional translator. Rask.AI requires a monthly paid subscription, whereas MultiVoice.AI offers payment based solely on usage (no subscription fee).
ElevenLabs
Language support: 29 languages.
Features: Preservation of the original voice and speech style, automatic speaker detection.
Advantages: Suitable for creators and educational content.
Dubformer
Language support: 130+ languages.
Features: Full-cycle dubbing with voice cloning, emotional delivery, and automatic mixing.
Advantages: Optimized for television content, including films and series.
Perso.ai
Language support: 32 languages.
Features: Multi-voice support, video synchronization, script editing, and access to AI avatars.
Advantages: Suitable for creating content with an individual style and visual elements.
Maestra.ai
Language support: 125+ languages.
Features: Automatic transcription, subtitle creation, and voice-over for video and audio files.
Advantages: A universal platform for various types of content.
Smartcat
Language support: 280+ languages.
Features: AI voice-over and automatic subtitles with editing capabilities in Smartcat Editor.
Advantages: Suitable for professional translators and localizers.
VMEG.ai
Language support: 170+ languages.
Features: Video synchronization, support for 7,000+ voices, accurate lip syncing.
Advantages: Suitable for large projects requiring high-quality localization.
CAMB.AI
Language support: 150+ languages.
Features: Real-time localization, including sports broadcasts.
Advantages: Suitable for broadcasts and live streams.
Advantages of Dubbing Studio:
- An intuitive graphical interface, which includes an English version.
- It's a client-server architecture. All file processing is handled on our end, using our GPU servers.
- It is possible to Dub the video Yourself (with the help of AI) or to use a professional translator from MultiVoice.AI.
- Voice-over ar Dubbing for a 90-minute video completed in 1-2 days.
- Works with YouTube, Google Disk, Tik Tok, Tencent, Vimeo, Twitch, Ruble, Spotify, Dailymotion.
Reviews
Our clients
Movify.CY Ltd.
Reg.nr. ΗΕ461401
Vat 60077645I
Agia Zoni, 30a, Fantaros Court,
Flat/Office 303, 3027, Limassol,
Cyprus.