speech
from google import genai from google.genai import types import wave # Set up the wave file to save the output: def wave_file(filename, pcm, channels=1, rate=24000, sample_width=2): with wave.open(filename, "wb") as wf: wf.setnchannels(channels) wf.setsampwidth(sample_width) wf.setframerate(rate) wf.writeframes(pcm) client = genai.Client() response = client.models.generate_content( model="gemini-2.5-flash-preview-tts", contents="Say cheerfully: Have a wonderful day!", config=types.GenerateContentConfig( response_modalities=["AUDIO"], speech_config=types.SpeechConfig( voice_config=types.VoiceConfig( prebuilt_voice_config=types.PrebuiltVoiceConfig( voice_name='Kore', ) ) ), ) ) data = response.candidates[0].content.parts[0].inline_data.data file_name='out.wav' wave_file(file_name, data) # Saves the file to current directory
