All models
OpenAI
GPT Audio
Voice conversation partner: hears your voice, answers by voice
This model is not in the catalog right now.
See available modelsAbout the model
GPT Audio works through an ordinary dialogue but understands voice as input and answers by voice or text, your choice. It is a conversational model, not a reader: it responds to the meaning of what was said, and when voicing a prepared text it may deviate from it word for word. If you need text spoken verbatim, take a speech model from the speech synthesis section. Available from Russia without a VPN, billed in rubles.
Strengths
- Voice in, voice out
- Answer by voice or text, your choice
- Works through a familiar dialogue request
Where to use it
- Voice assistants
- Voicing answers on the fly
- Understanding a dictated question