Loading this story…
Google's third audio release in under a month adds two speech models. Gemini 3.8 Flash TTS is for designed voices and directed dialogue. Flash-Lite TTS is the cheaper, higher-volume tier. Google says the pair took the top two spots on Hume AI's overall quality index.
Published · 2 sources
Confirmed that both models were released, that Google published the Hume scores, and that the developer surfaces are the API and AI Studio. That is Google's post, with the same launch reported by SiliconANGLE. A docs change that withdraws a surface, or a correction of the Hume figures, would be what would move this entry.
If you generate narration, dialogue, or in-product speech, this is a live API, not a demo. Flash TTS is the one Google points at character design and line-by-line direction. Flash-Lite TTS is the one it points at dubbing and voice agents where cost and volume matter.
Google's own rollout note is narrower than "available everywhere." Developers get the Gemini API and Google AI Studio now. Gemini Notebook is in the consumer path. Gemini Enterprise API access is listed as coming soon.
blog.google
Primary source
Open original (opens in a new tab)SiliconANGLE
siliconangle.com
Secondary
Open original (opens in a new tab)Learn the idea behind this story
Directed speech is a prompting problem: the line and the delivery instruction are separate, which is the habit the lab trains.
Open the prompt lab