Pay for what you use. See what the engine costs.

For every call you see where the money went: voice model, text model, transcription and speech synthesis on separate lines. Below are the raw list prices of the providers behind each engine, as examples.

Example prices. These are the providers' list prices as of November 2026 and they change from provider to provider. They are not our selling price. Costs in your call records use the unit prices configured at the time of the call.

GPT-Live 1

gpt-live-1 · gpt-5.6-luna

ItemUnitPrice (USD)
Voice model (gpt-live-1)
Billed per second
per minute $0.05
Background model, input per 1M tokens $0.20
Background model, cached input per 1M tokens $0.02
Background model, output per 1M tokens $1.20

About $0.05 per minute, plus the background model's tokens.

Gemini Live

gemini-3.8-live

ItemUnitPrice (USD)
Audio input
≈ $0.005/min
per 1M tokens $3.00
Audio output
≈ $0.018/min
per 1M tokens $12.00
Text input per 1M tokens $0.75
Text output per 1M tokens $4.50

Example: a minute with continuous audio in and out comes to about $0.023.

OpenAI Realtime

gpt-realtime-2.1-mini

ItemUnitPrice (USD)
Audio input per 1M tokens $10.00
Cached audio input per 1M tokens $0.30
Audio output per 1M tokens $20.00

Billed by audio tokens. Transcription with gpt-4o-mini-transcribe is charged separately.

Classic

Nova-3 · gpt-4.1-mini · gpt-4o-mini-tts

ItemUnitPrice (USD)
Transcription (Deepgram Nova-3) per minute $0.0077
LLM input (gpt-4.1-mini) per 1M tokens $0.40
LLM cached input per 1M tokens $0.10
LLM output per 1M tokens $1.60
Speech, text input (gpt-4o-mini-tts) per 1M tokens $0.60
Speech, audio output per 1M tokens $12.00

Every stage shows up on its own line.

Platform fee

We are onboarding customers one by one for now, and the platform fee depends on volume and setup. Tell us how many calls you expect and which phone line you use, and we will send you a quote.

Contact us

info@hubvocal.com

Telephony

Numbers and phone minutes are charged by your carrier (DIDWW, Twilio) or your own SIP trunk. They are not part of the engine costs above.

How you keep track

  • A cost breakdown per provider in every call record
  • A monthly spending limit per organization; new calls stop once it is reached
  • Usage by provider and by day, in the dashboard and through GET /v1/usage