- speech
- $15.00
- Released
- 2w ago
Cost/1M characters
AssemblyAI-TTS-1 is a text-to-speech model from AssemblyAI, announced as generally available. The launch announcement leads with natural-sounding speech at very low latency, claiming audio streams within 90 ms, and says the model was trained on millions of hours of audio, including real conversational data. It is built for voice agents, where the maker argues that natural, fast speech keeps callers from hanging up on automated phone systems.
How it compares
Frequently asked questions
Quality and benchmarks
The closest alternatives to TTS 1 are Simba 3.5 by Speechify, HeyGen Voice by HeyGen, and MAI-Voice-2.1-Flash by Microsoft. They were matched on model kind and release date, with no comparable price.
Simba 3.5TTS 1 vs Simba 3.5HeyGen VoiceTTS 1 vs HeyGen VoiceMAI-Voice-2.1-FlashTTS 1 vs MAI-Voice-2.1-Flash
Licensing and access
No, TTS 1 is closed, so it is neither open source nor open weights.
Basics
TTS 1 was made by AssemblyAI and released on October 27, 2026.
No, TTS 1 is closed and costs $15.00 per 1M characters for speech.
Media
TTS 1 costs $15.00 per 1M characters for speech.
TTS 1 supports the voices aai-antonioluis, aai-audrey, aai-carrie, aai-constanza, aai-donatien, aai-ernest, aai-ewald, aai-gunther, aai-hailey, aai-jorg, aai-kyla, aai-magdalena, aai-malvina, aai-nancy, aai-oriol, aai-peter, aai-sara, aai-steven, aai-titus, aai-troy.
Generated from Vertical Tabs data, updated November 12, 2026.
Discussion
No comments yet
Be the first to share your thoughts.