On September 10, 2026, OpenAI introduced GPT‑Live‑1 in the API—a Product Release for natural, full-duplex voice conversations with stronger instruction following, custom voices, and telephony support. The model docs list the API id as gpt-live-1. Those figures are USD as published. CAD is on the invoice. This desk is not inventing a Canadian list price.
gpt-live-1. Full-duplex voice: listen and speak at once.What OpenAI listed for GPT-Live-1
The product post frames GPT‑Live‑1 as bringing ChatGPT’s natural, full-duplex conversations to the API, with more control over how voice agents speak and act—tone, pace, and style via system prompt, custom voices, silent context / background-noise handling, and long-session reliability. The docs page prices voice sessions at $0.05 per minute, billed per second, and states session duration is not rounded up to the next whole minute. Backend Responses calls use normal pricing for the configured model and tools.
OpenAI reports Speak early evaluations cut interruptions by almost 80% versus previous turn-based systems, and Full Duplex Bench improves by 30 percentage points over GPT‑Realtime‑2.1. Concurrent session rate limits by tier are on the docs table (Free not supported; Tier 1 = 25 … Tier 5 = 500).
“Compared to our cascaded build, GPT‑Live‑1 simplified our code base by 80% and removed 23K lines of code. This enabled natural, real-time patient conversations & freed our team to improve the experience from booking an appointment to navigating care.”Tony Stoyanov, Co-Founder & CTO, as published on OpenAI’s GPT‑Live‑1 product post, September 10, 2026
What it costs
These are OpenAI’s published USD rates for the front-end voice layer. Do not quote them as CAD. CAD is on the invoice. Backend model and tool usage are separate line items.
| Line | Rate | Billing | Notes |
|---|---|---|---|
| Voice sessions | $0.05 / min | Per second | Not rounded up to whole minute |
| Backend model | Separate | Model rates | e.g. GPT-6 Astra or third-party |
| Tool usage | Separate | Tool rates | Billed apart from voice layer |
| Telephony | Supported | — | Phone-call voice agents |
What Alberta operators should do
Alberta work here means AI chatbot lead generation and AI automation Alberta—phone-line receptionist, missed-call capture, and booking flows for shops that lose leads when nobody picks up. GPT-Live-1’s telephony support is the shop path for a voice front door that can listen and speak at the same time, then hand hard reasoning to a backend text model.
Foreign hosted API is not a private Alberta path. For regulated data—client health notes, tenders, credentials—prefer a private AI security stack with a human review gate before anything leaves the shop. If the workflow has to stay on the desk with clear ownership, that is AI consulting work, not a public paste into a third-country endpoint. Quote USD as published; expect CAD on the invoice.
The guardrail
Price is not permission. A $0.05-per-minute voice card does not put inference in Alberta, and it does not authorize shipping regulated client audio offshore without a written data boundary. For AI chatbot lead generation, use GPT-Live-1 when the job fits a full-duplex phone agent, keep a human in the loop on bookings and care-adjacent intents, and keep sensitive workloads on a private agents path.
gpt-live-1) as a full-duplex voice front door for Alberta AI chatbot lead generation and missed-call capture—$0.05/min voice layer, billed per second, backend billed separate, telephony supported. USD as published; CAD on the invoice. Foreign API ≠ private/on-prem; put a human review gate on bookings and regulated intents, and prefer private agents for client data. Pair the voice layer with a backend you control, not a paste into a nameless endpoint.