Answer quests about audios || Responder perguntas referentes à um áudio

How can I pass an audio to gpt for analyzing, transcribing and answering some questions about the audio, such as a conversation between two people, and for gpt to divide the conversation into actors and their respective speeches and provide an explanation about the content of the conversation.

Como eu posso passar um áudio para o gpt analisar, transcrever e responder perguntas sobre? Exemplo, passo uma conversação entre duas pessoas e solicito para que ele separe em atores e suas falas e ele o faz, por fim ele cria uma breve explicação do teor da conversa, isto é\seria possível?

Hey there and welcome to the community!

You would not be passing it to GPT, but rather the Transcriptions API (Or the whisper API if you’re feeling fancy).

Once the Transcriptions API turns the speech into text, you can feed that data to GPT-4 to analyze the conversation further.

Note that it is not possible to analyze the raw audio data, only the test that’s produced afterwards.