Vertex AI 텍스트 음성 변환

Vertex AI 텍스트 음성 변환 (Text to Speech)

Chirp3 HD 음성과 Gemini TTS를 사용하는 Google Cloud Text-to-Speech를 LiteLLM에서 사용하는 방법을 알아봐요.

출처: 문서

본문

Chirp3 HD 음성

고품질 Chirp3 HD 음성을 갖춘 Google Cloud Text-to-Speech API.

빠른 시작

LiteLLM Python SDK

from litellm import speech
from pathlib import Path

speech_file_path = Path(__file__).parent / "speech.mp3"
response = speech(
    model="vertex_ai/chirp",
    voice="alloy",  # OpenAI voice name - automatically mapped
    input="Hello, this is Vertex AI Text to Speech",
    vertex_project="your-project-id",
    vertex_location="us-central1",
)
response.stream_to_file(speech_file_path)

LiteLLM AI Gateway

1. config.yaml 설정:

model_list:
  - model_name: vertex-tts
    litellm_params:
      model: vertex_ai/chirp
      vertex_project: "your-project-id"
      vertex_location: "us-central1"
      vertex_credentials: "/path/to/service_account.json"

2. Proxy 시작:

litellm --config /path/to/config.yaml

3. 요청하기:

curl http://0.0.0.0:4000/v1/audio/speech \
  -H "Authorization: Bearer ***" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vertex-tts",
    "voice": "alloy",
    "input": "Hello, this is Vertex AI Text to Speech"
  }' \
  --output speech.mp3

import openai

client = openai.OpenAI(api_key="sk-", base_url="http://0.0.0.0:4000")

response = client.audio.speech.create(
    model="vertex-tts",
    voice="alloy",
    input="Hello, this is Vertex AI Text to Speech",
)
response.stream_to_file("speech.mp3")

음성 매핑

LiteLLM은 OpenAI 음성 이름을 Google Cloud 음성으로 매핑해요. OpenAI 음성 또는 Google Cloud 음성을 직접 사용할 수 있어요.

Google Cloud 음성 직접 사용

from litellm import speech

# Pass Chirp3 HD voice name directly
response = speech(
    model="vertex_ai/chirp",
    voice="en-US-Chirp3-HD-Charon",
    input="Hello with a Chirp3 HD voice",
    vertex_project="your-project-id",
)
response.stream_to_file("speech.mp3")

Dict로 전달 (다국어):

from litellm import speech

# Pass as dict for full control over language and voice
response = speech(
    model="vertex_ai/chirp",
    voice={
        "languageCode": "de-DE",
        "name": "de-DE-Chirp3-HD-Charon",
    },
    input="Hallo, dies ist ein Test",
    vertex_project="your-project-id",
)
response.stream_to_file("speech.mp3")

AI Gateway:

curl http://0.0.0.0:4000/v1/audio/speech \
  -H "Authorization: Bearer ***" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "vertex-tts",
    "voice": {"languageCode": "de-DE", "name": "de-DE-Chirp3-HD-Charon"},
    "input": "Hallo, dies ist ein Test"
  }' \
  --output speech.mp3

사용 가능한 음성은 Google Cloud Text-to-Speech Console에서 찾아볼 수 있어요.

원시 SSML 전달

입력에 <speak> 태그가 있으면 LiteLLM이 자동으로 SSML로 감지하고 그대로 전달해요.

from litellm import speech

ssml = """
<speak>
    Hello, world!
    <break time="1s"/>
    This is a test of the <say-as interpret-as="characters">text-to-speech</say-as> API.
</speak>
"""

response = speech(
    model="vertex_ai/chirp",
    voice="en-US-Studio-O",
    input=ssml,  # Auto-detected as SSML
    vertex_project="your-project-id",
)
response.stream_to_file("speech.mp3")

SSML 모드 강제:

from litellm import speech

# Force SSML mode with use_ssml=True
response = speech(
    model="vertex_ai/chirp",
    voice="en-US-Studio-O",
    input="Speaking slowly",
    use_ssml=True,
    vertex_project="your-project-id",
)
response.stream_to_file("speech.mp3")

지원 파라미터

비동기 사용법

import asyncio
from litellm import aspeech

async def main():
    response = await aspeech(
        model="vertex_ai/chirp",
        voice="alloy",
        input="Hello from async",
        vertex_project="your-project-id",
    )
    response.stream_to_file("speech.mp3")

asyncio.run(main())

Gemini TTS

오디오 출력 능력을 갖춘 Gemini 모델을 chat completions API로 사용해요.

warning 제한 사항:

• pcm16 오디오 형식만 지원
• 스트리밍은 아직 미지원
• modalities: ["audio"] 설정 필수
• LiteLLM Proxy에서 사용할 때는 오디오 파라미터를 활성화하기 위해 요청 본문에 "allowed_openai_params": ["audio", "modalities"]를 포함해야 함

빠른 시작

from litellm import completion
import json

# Load credentials
with open('path/to/service_account.json', 'r') as file:
    vertex_credentials = json.dumps(json.load(file))

response = completion(
    model="vertex_ai/gemini-2.5-flash-preview-tts",
    messages=[{"role": "user", "content": "Say hello in a friendly voice"}],
    modalities=["audio"],
    audio={
        "voice": "Kore",
        "format": "pcm16"
    },
    vertex_credentials=vertex_credentials
)
print(response)

AI Gateway:

model_list:
  - model_name: gemini-tts
    litellm_params:
      model: vertex_ai/gemini-2.5-flash-preview-tts
      vertex_project: "your-project-id"
      vertex_location: "us-central1"
      vertex_credentials: "/path/to/service_account.json"

curl http://0.0.0.0:4000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer ***" \
  -d '{
    "model": "gemini-tts",
    "messages": [{"role": "user", "content": "Say hello in a friendly voice"}],
    "modalities": ["audio"],
    "audio": {"voice": "Kore", "format": "pcm16"},
    "allowed_openai_params": ["audio", "modalities"]
  }'

지원 모델

• vertex_ai/gemini-2.5-flash-preview-tts
• vertex_ai/gemini-2.5-pro-preview-tts

사용 가능한 음성은 Gemini TTS 문서를 참고해요.

고급 사용법

from litellm import completion

response = completion(
    model="vertex_ai/gemini-2.5-pro-preview-tts",
    messages=[
        {"role": "system", "content": "You are a helpful assistant that speaks clearly."},
        {"role": "user", "content": "Explain quantum computing in simple terms"}
    ],
    modalities=["audio"],
    audio={"voice": "Charon", "format": "pcm16"},
    temperature=0.7,
    max_tokens=150,
    vertex_credentials=vertex_credentials
)

더 알아보기 (Learn more)

• Google Cloud Text-to-Speech 문서
• Gemini TTS 문서