Voice on ToRun is not one feature — it is three, and each one picks from multiple providers.
Realtime conversations
Live voice mode connects you to speech-native models for real back-and-forth conversation — interruptions, tone, and all. Sessions are metered by audio time with the price shown before you connect, and idle sessions are wound down automatically so a forgotten tab cannot quietly bill you.
Text-to-speech
TTS spans providers and voice rosters — from fast, inexpensive voices for utility work to expressive ones for narration. The voice picker shows the per-character or per-minute price next to each option, so the cost of "make it sound nicer" is a number, not a surprise.
Transcription
Speech-to-text routes to fast inference providers when speed matters and to higher-accuracy models when fidelity matters. Upload audio or speak directly; the transcript lands in your chat where you can immediately summarize, translate, or act on it.
One bill, one ledger
All three run through the same billing pipeline as text: one record per call, priced at the moment of use, visible in your ledger seconds later. Voice is a first-class citizen, not a bolted-on demo.