a live conversation can’t wait on a far GPU
Bring your recognition, language and synthesis models. YOM AI runs them in live sessions near the speaker; heavy turns go to a model you choose.
only the heavy turn leaves the node
Two hops at most. The light work stays on the machine near the speaker; the heavy turn goes out to an outside model you choose and comes back into the same session.
-
the person speaks
the browser or the app, no download and no plugin
-
the node hears
your recognition model, on a GPU near the speaker
-
the node thinks
light turns here; a heavy one hops out and comes back
-
the node speaks
your synthesis model, back down the same connection
what it is
- your own models (bring your own model), run together in one session on a GPU near the speaker
- a stateful connection that stays open for the whole conversation
- heavy turns routed to an outside model you choose, and back
- metered per session, like a game, and settled when it ends
built for
- companions and characters that must stay in persona, on models you own
- tutors with their own speech and pronunciation models
- games and worlds with on-brand voices for their characters
Consumer nodes run best-effort, without a guaranteed service level. Regulated and safety-critical calls belong on operator-grade capacity, which we scope with you.
a conversation costs what a game costs
A voice session is metered like a game session: up to an hour, settled when it ends. Tell us which models you run and we quote it.
from $0.04
one conversation is one session, up to an hour, to $0.24 in the top compute class
how every session here is pricedwho is already asking
-
voice teams that own their models
built for companions, tutors and character voices on models their teams own
the ask
bring the conversation
Tell us which models you run, where your speakers are and how many conversations you carry. We come back with where the network can place them today and what it costs.
bring your voice models