What VoiceThere is
What VoiceThere sells
VoiceThere sells hosted real-time voice agents, and voice and data apps, over WebRTC. You write the agent in TypeScript with @voicethere/agent, deploy it with @voicethere/cli, and connect people from a browser or a Node client. We run the signaling, the sessions, the speech workers and the scaling, so you do not operate a WebRTC stack yourself.
A session can carry voice, data channels or both. The same platform runs a voice support agent, a multiplayer game or a shared whiteboard.
What you get
- Voice sessions over WebRTC for browser and Node clients, with barge-in, so callers can interrupt the agent.
- Built-in speech recognition and synthesis on the worker, or your own keys for vendors such as OpenAI, Deepgram or ElevenLabs. See voice vendors.
- LLM voice agents you edit in the dashboard: provider, model, prompt and tools. See LLM agents.
- Data-only and multiplayer sessions, with an optional project Redis for shared state.
- A dashboard and REST API for projects, builds, promote and rollback, plus agent logs and session error tracking.
Who it is for
Developers who want to ship a voice or realtime feature without building the media, signaling and scaling layers first, and teams that want to deploy the same agent bundle they run locally.
How billing works
You pay for what a project uses, in credits. Each plan includes monthly credits, and you can buy credit packs. New sessions need available credits on the project. Current prices are on the pricing page.
Where to go next
- Quickstart: your first agent, live.
- Pricing: plans and credit packs.
- Voice vendors: which speech providers you can use.