Models/shipped/2026-09-20

Google Introduces Gemini 3.8 Text-to-Speech Models and Related AI Features

Google has announced the release of new Gemini 3.8 models, with a particular focus on text-to-speech (TTS) capabilities. The company's official blogs confirm the introduction of Gemini 3.8 TTS, highlighting its ability to generate expressive, multilingual voices. Reports indicate these models, including "Flash" and "Flash-Lite" versions, are available via API and AI Studio. Google claims features such as custom voice creation in over 100 languages and improved agent functionalities like thinking levels and tool retries. While the launch is confirmed by Google, the full extent of the models' capabilities and performance are based on the company's own statements.

20 articles from 18 outlets covered this story. Their coverage differs on 4 points. The underlying claim is sourced from a shipped.

What do all outlets agree on?

18 outlets covered “Google Introduces Gemini 3.8 Text-to-Speech Models and Related AI Features”. All of them report the following:

Google launched new Gemini 3.8 models
The release includes text-to-speech (TTS) capabilities
Models are available via API and AI Studio
The models support multilingual voices
"Flash" and "Flash-Lite" versions are part of the release

Did outlets disagree about this?

Yes. Coverage of “Google Introduces Gemini 3.8 Text-to-Speech Models and Related AI Features” differs on 4 points. Each account below is how a different outlet described the same event:

The primary focus of the Gemini 3.8 release is text-to-speech.

Google DeepMind Blog, blog.google, Crypto Briefing, Unite.AI, Seeking Alpha, NDTV Profit

The Gemini 3.8 release also significantly includes or enables broader multimodal AI and agentic capabilities.

Dynamic Business, SitePoint, The Sequence

The release of Gemini 3.8 models raises AI safety concerns.

Pluang

The release is a standard product update, without specific mention of safety concerns.

Google DeepMind Blog, blog.google, Dynamic Business, SitePoint, Crypto Briefing, Unite.AI, Seeking Alpha, MediaPost, NDTV Profit, The Sequence

Which figures do outlets report differently?

1 figure in this story is reported with conflicting values:

Number of languages supported for custom voice creation100+ Languages (NDTV Profit) vs multilingual (Crypto Briefing, Google DeepMind Blog, blog.google, Unite.AI, Seeking Alpha, Dynamic Business, Pluang, SitePoint, The Sequence, MediaPost)