Providing Voice Dialogue Technology.

We provide the voice-dialogue technology we research and develop in-house as APIs and frameworks.

Voice Dialogue API (Cloud)

A streaming voice-dialogue API in which characters speak in voices designed from a written description.

This demo runs in the cloud, and the results include network latency. The demo uses servers overseas, which adds latency; if you need lower latency, a Tokyo region is also available.

Folding prompts and tool results into speech

Prompts and tool results that arrive while the AI is speaking are folded naturally into the dialogue, without stopping the speech.

On-device voice dialogue

We also offer features that support on-device voice dialogue. For example, low-latency speech synthesis like the demo below, and a framework that includes low-latency speech recognition.