AssemblyAI launches first TTS model: 80ms TTFB, 4,000+ voices, $15/1M chars
ycombinator · x · 2026-10-28
AssemblyAI has shipped AssemblyAI-TTS-1, its first text-to-speech model, claiming no other TTS model beats it on the combination of accuracy, naturalness, and cost.
Key specs: 80ms time-to-first-byte, instant voice cloning, 4,000+ voices, unlimited concurrency, and $15 per 1M characters — a frontier-quality, aggressive-pricing option for large-scale speech synthesis.
More from Multimodal
- VEED launches dubbing API with built-in lip sync that preserves emotion and accent — hey_abusiddik · 2026-10-28
- Indie dev vibe-codes Try2Create, a one-click ComfyUI frontend for image, video and audio generation — ItsBalanse · 2026-10-28
- MiniMax H3 demo generates a continuous music video with seamless audio-video sync — EinhornArt · 2026-10-28
- Mundo turns photos of real places into explorable, editable interactive worlds — Scobleizer · 2026-10-28
- Google's new AI world model reportedly simulates entire worlds in real time — heykahn · 2026-10-28
- NVIDIA open-sources Nemotron 3.5 Transcribe 30B ASR model; fine-tuning cuts word error rate 45.8% — NVIDIAAI · 2026-10-28