speech_to_text.py

from openai import OpenAI
import os
client = OpenAI(
    api_key=os.getenv("OPENAI_API_KEY"),
    organization=os.getenv("OPENAI_ORG_ID"),  # 可选 OpenAI API 中的组织 ID
    timeout=30.0  # 默认超时时间
)

def transcribe_audio():
    audio_file = open("speech.mp3", "rb")
    transcript = client.audio.transcriptions.create(
        model="whisper-1",
        file=audio_file,
        language="zh",           # 音频语言
        prompt="这是一段中文演讲", # 提示词以提高准确性
        response_format="text",  # text, json, srt, verbose_json, vtt
        temperature=0.2,        # 控制随机性
        timestamp_granularities=["word", "segment"]  # 时间戳粒度
    )
    return transcript.text