Models by OpenAI
OpenAI in Tõlk: the default live interpreter.
OpenAI’s realtime translation model is the engine most Tõlk events start on. Its recognition, translation and voice models also power Dictionary-assisted routes and Studio.
- Speech-to-speech
- Speech-to-text
- Text translation
- Text-to-speech
Models and how they run.
- Speech-to-speech
Realtime translation
gpt-realtime-translate
One live session per language, interpreter style.
- Speech-to-text
Realtime transcription
gpt-4o-mini-transcribe
Default stage recognition: captions, transcripts, summaries.
- Speech-to-text
Realtime transcription (full)
gpt-4o-transcribe
Larger recognition model, chosen in event settings.
- Text translation
Text translation
gpt-6-luna
Translates each phrase with your event Dictionary.
- Text-to-speech
Speech
gpt-4o-mini-tts
Reads translations live, and Studio’s Standard voices.
Why teams pick it
- Fast, natural interpreter-style delivery
- Selectable voices and speech speed
- Original-language captions from a separate recognition stream
- Full speech → text → speech route with your terms
Good to know
- The default engine, with Gemini and Qwen ready as backups.
- Organizers set the voice and speech speed per language.
Languages
Interpreter languages, and many more voices
Realtime translation speaks the world’s major languages directly. OpenAI voices read translations in far more on the speech → text → speech route.
Languages listeners can hear the translation in.
Realtime translationSpeech-to-speech13 languages
- Chinese
- English
- French
- German
- Hindi
- Indonesian
- Italian
- Japanese
- Korean
- Portuguese
- Russian
- Spanish
- Vietnamese
OpenAI voiceSpeech → text → speech59 languages
- Afrikaans
- Arabic
- Armenian
- Azerbaijani
- Belarusian
- Bulgarian
- Catalan
- Chinese
- Chinese (Simplified)
- Chinese (Traditional)
- Croatian
- Czech
- Danish
- Dutch
- English
- Estonian
- Filipino
- Finnish
- French
- Galician
- German
- Greek
- Hebrew
- Hindi
- Hungarian
- Icelandic
- Indonesian
- Italian
- Japanese
- Kannada
- Kazakh
- Korean
- Latvian
- Lithuanian
- Macedonian
- Malay
- Marathi
- Nepali
- Norwegian
- Norwegian (Bokmål)
- Persian
- Polish
- Portuguese
- Portuguese (Brazil)
- Portuguese (Portugal)
- Romanian
- Russian
- Serbian
- Slovak
- Slovenian
- Spanish
- Swahili
- Swedish
- Tamil
- Thai
- Turkish
- Ukrainian
- Urdu
- Vietnamese
Tõlk shows which engines can serve each language when you set up an event.
Where OpenAI works best.
No setup on your side
OpenAI runs through Tõlk’s own accounts. You never connect a key, and if an engine has trouble, affected languages switch to a backup and back again on their own.
OpenAI is a trademark of OpenAI. Tõlk uses its models through OpenAI’s public APIs; this is not a partnership or endorsement.
More of the mix
All routes and models
Another model in mind? We’ll bring it in.
- Another providerA vendor or model you already use or trust.
- Open-source modelsOpen-weights speech, translation or voice models.
- Self-hostedModels on your own servers or private cloud.
The Live AI Translation Hub
Ready for your next moment?
Tell us what to make multilingual. We pilot it on your content, then you decide.