Frontier Flux Original
Google launches Gemini 3.8 Flash TTS and Flash-Lite TTS
On 23 September 2026 Google launched Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, two text-to-speech models rolling out on the Gemini API, Google AI Studio, Gemini Notebook, and Google Vids. Hume Voice Design ranks in the launch post stay Google-reported. Enterprise API is still coming soon.
Abstract concept art for a two-SKU text-to-speech launch. AI-generated.
Frontier Flux / AI-generated
Google on 23 September 2026 launched two text-to-speech models in the Gemini 3.8 Audio family: Gemini 3.8 Flash TTS for voice design and line-by-line direction, and Gemini 3.8 Flash-Lite TTS for high-volume generation. The launch post says they complement Gemini 3.5 Live Translate, 3.5 Transcribe, 3.8 Live, and 3.8 Live Extended Thinking.
From the 23 September 2026 @GoogleDeepMind post:
"Create and deploy custom audio with our new text-to-speech models: Gemini 3.8 Flash TTS: Design unique voices with distinct accents and characteristics. Gemini 3.8 Flash-Lite TTS: Built for efficiency and scale, choose from your created styles or our expansive…"
The public embed cuts the post there.
The Google launch post (Leland Rechis and Alan Cowen) says Flash TTS is rolling out starting today in the Gemini API, Google AI Studio, and Gemini Notebook, with Gemini Enterprise "coming soon." Flash-Lite TTS is rolling out in the Gemini API, Google AI Studio, and Google Vids, also with Enterprise coming soon. Developer docs name stable ids `gemini-3.8-flash-tts` and `gemini-3.8-flash-lite-tts`, and list Flash-Lite as the recommended replacement for `gemini-3.1-flash-tts-preview`. The Gemini 3.8 Audio model card lists the TTS pair as text in up to 8K tokens and audio out at 64K tokens, based on Gemini 3 Pro. DeepMind's speech generation page marks both as general availability.
Google-reported evals in the launch post: Flash TTS first overall on Hume AI's Voice Design Benchmark at 71.4 and first in accent modeling at 60.8; Flash TTS and Flash-Lite TTS first and second on Hume's Overall Quality Index. Google also cites Voice Arena "top positions" in Japanese, Brazilian Portuguese, Vietnamese, Modern Standard Arabic, Mexican Spanish, and Hindi. Developer docs list 130 languages for Flash TTS and 101 for Flash-Lite TTS.
The launch post attributes generative voice design, a 2,000-plus voice library, and 30-second replication (consent verification, SynthID watermarking, C2PA credentials) to Flash TTS. Developer docs say both SKUs support prebuilt voices, voice design, and voice replication. Voice replication through AI Studio is not available in Illinois, Texas, the EEA, UK, Switzerland, and India.
Key facts
- Social alert: @GoogleDeepMind on X, 23 September 2026
- Official primary: Gemini 3.8 text-to-speech says hello
- API ids (Google): `gemini-3.8-flash-tts`, `gemini-3.8-flash-lite-tts`
- Model card: Gemini 3 Pro base; 8K text in, 64K token audio out
How to read
Google is rolling out a two-SKU TTS stack. Hume scores and Voice Arena placements are Google claims. The speech page marks general availability; the blog says rolling out starting today. Enterprise access is still coming soon.
Independence
Frontier Flux is not affiliated with Google or DeepMind. The X post is the alert. Rankings here are Google-reported, not our ranking.