Google unveiled Gemini 3.8 Flash TTS on September 23, 2026. Just describe the voice you want in text, clone a voice from a 30-second sample, and choose from 2,000+ presets across 100+ languages. The paradigm for AI voice production is shifting.
Google officially launched Gemini 3.5 Transcribe on August 26, 2026. It achieves an average Word Error Rate (WER) of 2.6% across 85+ languages, supporting both non-streaming and real-time streaming modes. With speaker diarization, word-level timestamps, and up to 1,000 custom vocabulary terms, this model is set to transform education and content creation workflows.
The Gemini 2.5 Flash/Pro TTS API brings 30 HD voices and 24 languages to educational audio content creation. Emotion control from cheerful to somber, multi-speaker dialogue generation, and context-aware pacing β exploring what this API makes possible from an EdTech perspective.