Google's Gemini 3.8 Flash TTS is out; its API price doubles in 2027

Google made two text-to-speech models generally available on the Gemini API on September 22, 2026: Gemini 3.8 Flash TTS (gemini-3.8-flash-tts) and Gemini 3.8 Flash-Lite TTS (gemini-3.8-flash-lite-tts). Both launch at a price that runs through December 31, 2026. From January 1, 2027, the pricing page shows every rate doubling.

The two models

• Gemini 3.8 Flash TTS is the flagship. Google describes it as built for studio-grade voices, nuanced acting and regional dialects, and stable over long, multi-turn output. The models page says it covers 130 languages.
• Gemini 3.8 Flash-Lite TTS is the cheaper one, built to replace gemini-3.1-flash-tts-preview for high-volume use such as real-time voice agents. It covers 101 languages.

Google also launched a Voices endpoint (/v1beta/voices) on the same day. It lets developers design lasting custom voices from a text prompt, copy a voice with consent verification, and search a library of more than 150 built-in and custom voices.

What it costs

Paid Standard tier, per 1M tokens, from Google’s pricing page on October 18, 2026:

Model	Text input, 2026	From Jan 1, 2027	Audio output, 2026	From Jan 1, 2027
3.8 Flash TTS	$0.50	$1.00	$9.00	$18.00
3.8 Flash-Lite TTS	$0.50	$1.00	$6.00	$12.00

Google converts this to $0.00225 per 10 seconds of audio for Flash TTS until the end of 2026 and $0.0045 from January 1, 2027. For Flash-Lite TTS it is $0.0015, then $0.003. Batch and Flex requests cost half the Standard rate. Both models are free of charge on the free tier.

For comparison, the preview model that Flash-Lite replaces, gemini-3.1-flash-tts-preview, lists $1.00 for text input and $20.00 for audio output per 1M tokens.

What to do

If you are building a budget for a voice product that runs into 2027, use the January 1, 2027 prices, not the launch prices. The cost per 10 seconds of speech doubles on that date.