OpenAI 推出更低成本的语音模型API ：gpt-4o-audio-preview （附代码）

from flask import Flask, request, jsonify, url_forimport base64import openaiimport osimport uuid
app = Flask(__name__)
# 配置OpenAI的API密钥openai.api_key = "在这里填写OpenAI_API_KEY"
# 检查并创建 static 目录if not os.path.exists('static'):    os.makedirs('static')
# 存储对话记录conversation_history = []
@app.route('/')def index():    return '''        <h1>GPT-4 Audio Web Application</h1>        <form id="questionForm">            <label for="question">Enter your question:</label>            <input type="text" id="question" name="question">            <button type="submit">Submit</button>        </form>        <div id="audioPlayer">            <h3>Response Audio:</h3>            <ul id="audioList"></ul>        </div>        <script>            document.getElementById('questionForm').addEventListener('submit', function(event) {                event.preventDefault();                let formData = new FormData(this);                fetch('/ask', {                    method: 'POST',                    body: formData                }).then(response => response.json())                .then(data => {                    if (data.conversation_history) {                        let audioList = document.getElementById('audioList');                        audioList.innerHTML = '';  // 清空列表                        data.conversation_history.forEach(item => {                            let listItem = document.createElement('li');                            let audioElement = document.createElement('audio');                            audioElement.src = item.audio_url;                            audioElement.controls = true;                            listItem.textContent = item.question;                            listItem.appendChild(audioElement);                            audioList.appendChild(listItem);                        });                    } else {                        alert('Error: ' + data.error);                    }                });            });        </script>    '''
@app.route('/ask', methods=['POST'])def ask_gpt():    try:        # 获取用户输入的问题        question = request.form['question']
        # 使用OpenAI API        client = openai        completion = client.ChatCompletion.create(            model="gpt-4o-audio-preview",            modalities=["text", "audio"],            audio={"voice": "alloy", "format": "wav"},            messages=[                {                    "role": "user",                    "content": question                }            ]        )
        # 生成唯一的文件名        unique_filename = f"{uuid.uuid4()}.wav"        audio_file_path = os.path.join('static', unique_filename)
        # 提取音频数据并保存        wav_bytes = base64.b64decode(completion.choices[0].message['audio']['data'])        with open(audio_file_path, "wb") as f:            f.write(wav_bytes)
        # 生成音频文件的URL        audio_url = url_for('static', filename=unique_filename)
        # 保存到对话历史        conversation_history.append({            "question": question,            "audio_url": audio_url        })
        # 返回对话历史        return jsonify({"conversation_history": conversation_history})
    except Exception as e:        return jsonify({"error": str(e)}), 500
if __name__ == '__main__':    app.run(debug=True)

pip install Flask openai

python app.py