01
Why it is on the roadmap
Voice agents require more than transcription and synthesis: turn-taking, interruption handling, model orchestration, tool permissions and deterministic failure behavior must work as one system.
Audelyr product roadmap
Roadmap · not available
Voice Agents
Planned orchestration for speech, reasoning, tool use and interruption-aware conversations. No Audelyr Voice Agent runtime or public API is available today.
Capabilities
One interface.
Clear boundaries.
Designed as composable infrastructure, with Preview labels wherever production behavior is not yet available.
01
Agent runtime — planned
Documented, observable and designed to fit into your existing product architecture.
02
Tool policies — planned
Documented, observable and designed to fit into your existing product architecture.
03
Barge-in — planned
Documented, observable and designed to fit into your existing product architecture.
04
Combined usage metering — planned
Documented, observable and designed to fit into your existing product architecture.
02
What exists today
Audelyr currently exposes speech and speech-translation building blocks. Customers can compose these with their own application logic, but Audelyr does not provide an Agent Runtime endpoint.
03
Planned boundary
A future provider-independent agent runtime may coordinate STT, LLM reasoning, tools and TTS without exposing upstream provider structures.
04
No availability claim
This page documents direction only. There is no Preview key, production key, price, SLA or launch date for Voice Agents.
Early Access
Build the interface
before the infrastructure.
Explore the honest Preview, then tell us what production voice needs from Audelyr.