Capabilities
Realtime voice
Speak naturally and receive spoken responses for fast conversational work.
Dictation
Convert speech into text for long prompts, notes, or messages.
Tool use
Let voice requests trigger configured tools through the normal tool-choice controls.
Context
Keep voice interactions grounded in the current thread and selected resources.
Configuration
Voice capabilities depend on the providers configured in your environment. Add speech or realtime API keys where your deployment expects them, and addASSEMBLYAI_API_KEY when you want AssemblyAI-backed dictation.