This page may contain stale information. Last updated: 2026-09-24
title: “Text-to-Speech”
aliases: [“TTS”]
category: “ai”
created_at: “2026-07-28”
updated_at: “2026-09-24”
sources: [“2026-09-23-gemini-3-8-text-to-speech”, “2026-09-23-gemini-tts-voice-design-tnw”, “2026-07-28-fish-audio-52m-seed-ai-voice”]
Definition
Text-to-speech (TTS) converts written text into spoken audio using neural acoustic and vocoder models; modern systems add emotion, multilingual support, and natural-language voice controls.
Key Points
- 2026-09-23: gemini-38-flash-tts and Flash-Lite TTS — natural-language voice design, line-by-line performance direction, 2,000+ preset voices, consent-gated replication; SynthID on all output (2026-09-23-gemini-3-8-text-to-speech)
- GitHub star counts and ARR claims are common traction signals but often company-reported
- Enterprise TTS APIs compete on latency, cloning quality, and pricing
- Roadmaps increasingly include audio understanding and speech-to-speech