Build real-time voice agents on Together AI
Mar 12, 2026
Summary
This post announces the native hosting of Deepgram (STT) and Cartesia (TTS) models on Together AI infrastructure, expanding the model choice for voice agents. It emphasizes the architectural benefit of co-locating STT, LLM, and TTS components within the same cluster to achieve end-to-end pipeline latency under 500ms. This co-location reduces inter-vendor network hops, improving speed, flexibility (access to intermediate transcript/response text), and reliability (single API, auth, billing, metrics). The post highlights the production use case of Decagon running concierge voice agents on this stack.
More from the Speech-to-Text (STT) APIs trail
- How Together AI built the world’s fastest speech-to-text stack
- Violin: An open-source video translation skill that breaks language barriers
- Deepgram speech-to-text and voice models now available natively on Together AI
- How speech models fail where it matters the most and what to do about it
- Announcing the fastest inference for realtime voice AI agents