Google brings Gemini 3.8 Flash TTS and Flash-Lite TTS to general availability

Google has moved its next-generation text-to-speech models, Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, out of preview and into general availability on the Gemini API, alongside a new dedicated Voices endpoint.

What's new

The Gemini API changelog entry for September 22, 2026 announces the two models are now GA: "Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS generally available (GA): Released our next-generation text-to-speech (TTS) audio models and the Gemini API Voices endpoint (/v1beta/voices)."

Gemini 3.8 Flash TTS is the higher-fidelity of the two, aimed at studio-grade voice quality, nuanced vocal acting, regional dialects, and stability across long, multi-turn audio generation. Gemini 3.8 Flash-Lite TTS is the cost-efficient counterpart, built to replace the older gemini-3.1-flash-tts-preview model in high-throughput production settings where cost per request matters more than peak fidelity.

The release also adds voice design and voice replication capabilities, plus access to what Google describes as an Extended Voice Library of more than 150 prebuilt and custom voices, available through the new Voices endpoint.

Context

The TTS GA lands in the same window as several other Gemini API changes this month, including new access restrictions on the older Gemini 2.5 model family and an update to Google's Antigravity coding agent. Google has been steadily building out audio and multimodal endpoints alongside its core Gemini 3.x text and reasoning models, and a dedicated Voices endpoint signals it now treats speech generation as a first-class API surface rather than a feature bolted onto the main generation endpoint.

Text-to-speech is also an increasingly contested part of the AI stack, with dedicated voice AI vendors like ElevenLabs competing directly against the speech capabilities that foundation model providers such as Google, OpenAI, and Anthropic build into their own platforms.

Why it matters

Moving from preview to GA, with a cost-efficient Flash-Lite tier and a purpose-built Voices endpoint, tells developers Google now considers Gemini's TTS stack production-ready rather than experimental. The explicit replacement path for the older flash-tts-preview model also signals Google's intent to consolidate its audio lineup around the 3.8 generation, giving teams building voice products on Gemini a clearer signal about which models to build on going forward.