Gemini 3.8 Flash TTS

Gemini 3.8 Flash TTS (gemini-3.8-flash-tts) is Google's flagship text-to-speech model, available through Gemini Enterprise. It's built for studio-grade voice fidelity, expressive acting, authentic regional accents, and stable long-form, multi-turn audio.

Capabilities

  • Acoustic fidelity and acting nuance: Delivers a wide emotional range, natural cadence, and close adherence to turn-level style directions and inline vocal events such as <laugh>, <sigh>, and <short pause>.
  • Long-form, multi-turn stability: Keeps voice identity, timbre, volume, and room tone consistent across long dialogues and multi-minute narration.
  • Regional accents and pronunciation: Supports regional accents and minority dialects in 130 languages.
  • Voice options: Works with 30 prebuilt voices, more than 2,000 curated voices in the Extended Voice Library, voices that you create with Voice design, and voices that you replicate with Voice replication.

For features, code samples, and prompting guidance, see Generate speech with Gemini TTS.

When to use which TTS model

Both Gemini 3.8 TTS models share the same request schema and prompting format. Choose the model that fits your workload:

Feature or workload Gemini 3.8 Flash TTS (gemini-3.8-flash-tts) Gemini 3.8 Flash-Lite TTS (gemini-3.8-flash-lite-tts)
Primary strength Voice fidelity, acting nuance, and dialect coverage High throughput, low latency, and cost efficiency
Best use cases Audiobooks, studio narration, complex multi-speaker dialogue, frequent vocal-burst tags, difficult pronunciation, regional dialects High-volume production, real-time voice agent cascades, read-aloud features, voice replication, everyday single-speaker speech
Supported languages 130 languages 101 languages

Pricing

Model ID gemini-3.8-flash-tts
Modalities
Text Input only
Image Not supported
Audio Output only
Video Not supported
Token limits Input token limit 8,192
Output token limit 16,384
Capabilities
Consumption options
Supported regions

Model availability

  • Global: global
Versions
  • gemini-3.8-flash-tts
    • Launch stage: Preview
    • Release date: September 28, 2026
Supported languages 130 languages. See Supported languages.

Get started

The following example generates single-speaker speech and saves it as a WAV file:

Python

from google import genai

client = genai.Client(enterprise=True, project="PROJECT_ID", location="global")

response = client.models.generate_content(
    model="gemini-3.8-flash-tts",
    contents=[{
        "role": "user",
        "parts": [{
            "text": "Have a wonderful day!",
            "speech_metadata": {"style": "cheerful and friendly"},
        }],
    }],
    config={
        "response_modalities": ["AUDIO"],
        "speech_config": {"voice_config": {"voice": "Kore"}},
    },
)

# The SDK has already decoded the base64 audio, so inline_data.data is a
# complete WAV file by default.
with open("out.wav", "wb") as f:
    f.write(response.candidates[0].content.parts[0].inline_data.data)

REST

curl -X POST \
  -H "Authorization: Bearer $(gcloud auth print-access-token)" \
  -H "Content-Type: application/json" \
  https://aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/global/publishers/google/models/gemini-3.8-flash-tts:generateContent \
  -d '{
    "contents": [{
      "role": "user",
      "parts": [{
        "text": "Have a wonderful day!",
        "speechMetadata": {"style": "cheerful and friendly"}
      }]
    }],
    "generationConfig": {
      "responseModalities": ["AUDIO"],
      "speechConfig": {
        "voiceConfig": {"voice": "Kore"}
      }
    }
  }' | jq -r '.candidates[0].content.parts[0].inlineData.data' | base64 --decode > out.wav

If you use an earlier Gemini TTS model, see the migration guide.