Google expanded its Gemini Audio suite with the API-only release of Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS. Available through the Gemini API and Google AI Studio, the models offer prompt-based voice control, allowing developers to direct delivery and creative tone using natural language script cues.

Flash TTS focuses on character voices and creative control, while Flash-Lite TTS offers lower-cost, high-volume generation. The release introduces voice replication from a 30-second audio clip, requiring verified verbal consent from the speaker to prevent unauthorized cloning.

To address safety concerns, Google embedded invisible SynthID watermarks into all audio outputs and applied C2PA content credentials specifically to replicated voices, with enterprise access rolling out soon.

Why it matters

  • Enables dynamic, fine-grained voice control via text prompts without requiring extensive manual SSML or audio editing.

  • Introduces explicit consent verification workflows and digital watermarking standards for enterprise voice cloning.

Source: marktechpost.com