AI voice models: the wrap-up comparison
Prices rechecked 14 Nov. Hume's old TTS docs now redirect, Chatterbox's README calls the repo an archive, and the main API prices still match.
Topic · 16 stories
16 stories tagged voice-agents on AI Tamer, newest first.
Prices rechecked 14 Nov. Hume's old TTS docs now redirect, Chatterbox's README calls the repo an archive, and the main API prices still match.
Nuance Labs launched Emma on Nuplex 1.0, a full-duplex audiovisual model that listens as it speaks, with a beta API that bans cloning and adds watermarks.
Cartesia's Sonic 3.6 ranks 4th for TTS from $37 per million characters. Hume's Octave and EVI APIs shut on 13 Nov. Ranks, cloning rules and a migration plan.
Deepgram is acquiring Daily. Pipecat stays BSD-2 under a new foundation, and the two launch RCS, an alpha format for tone and delivery cues.
AssemblyAI-TTS-1 is GA for voice agents: six languages, 4,000+ voices, instant cloning and about 90 ms to first audio, with no concurrency fees.
Eleven Realtime v1 Mini and v1 Max are single speech-to-speech models for voice agents, priced at $0.08 a minute plus LLM tokens.
Voxtral 3 pairs new Mistral transcription and speech models with a LiveKit-based voice agent SDK, on Mistral's API or self-deployed.
Browser echo-cancellation constraints negotiate a track setting. Learn what supportedConstraints, ideal, exact, and getSettings reveal, then test the microphone route that users actually use.
Twilio buffers outbound media and returns matching marks after playback or clear. An AI voice agent needs that distinction to account for speech interrupted by a caller.
A spoken system can lose you in the middle of a sentence, and the listener is left not knowing whether the thing happened. Design for that uncertainty.
Network audio arrives asynchronously while an AudioWorklet must fill each render block on time. A bounded ring buffer and explicit silence policy turn missing samples into predictable behavior.
A spoken confirmation that only reassures, without repeating back what was heard, can hide a mistake. The fix is to make the confirmation carry the content.
AssemblyAI announced Universal-3.6 Pro Realtime on 2026-09-29—streaming STT for voice agents, model id universal-3-6-pro. Drop-in versus 3.5 Pro (still available). $0.45/hr; billed by WebSocket session duration.
Voice activity detection finds speech and silence; turn detection, endpointing and interruption rules decide what an agent should do with those signals.
A practical map of speech recognition, language models, speech synthesis, direct speech-to-speech models, streaming, and the latency measurements that matter.
Alibaba's Qwen team upgraded its ASR, TTS and Realtime voice models and added TTS-Next and ASR-Next, while a Model Studio notice cut realtime voice rates by up to about 88% from September 22.