Dessa launches full-duplex speech model trained only on human conversation

Dessa introduced Dessa, billed as the most authentic full-duplex speech model, trained from scratch only on human conversation audio without text or pretrained models. It converses in real time with emotion, pauses and natural cadence, laughing mid-sentence, raising volume, and finishing sentences, behaviors absent in text-based voice AI.

• Amp puts AI agents to work beyond code with OpenAI models2026-10-14
• OpenAI adds GPT‑Live‑1 voice to API2026-09-10