2 2
3> For the complete documentation index, see [llms.txt](/llms.txt). Markdown versions of documentation pages are available by appending `.md` to the page URL.3> For the complete documentation index, see [llms.txt](/llms.txt). Markdown versions of documentation pages are available by appending `.md` to the page URL.
4 4
5GPT-Live handles a spoken conversation while a backend agent looks up information, uses tools, and completes tasks. It can listen while speaking, a capability called **full duplex**. Sending work to the backend is called **delegation**: the conversation can continue while that work runs.5GPT-Live handles a spoken conversation while a backend agent looks up information, uses tools, and completes tasks. It can listen while speaking (**full duplex**). Sending work to the backend is called **delegation**.
6 6
7For example, a user can ask about an order, add a detail while the backend checks its status, and hear the result when it is ready. You choose the backend model or agent independently of the voice model; Realtime uses one model for speech, reasoning, and tool selection.7For example, a user can ask about an order and add a detail while the backend checks its status. GPT-Live can keep talking with the user and explain the result when it arrives. You choose the backend model or agent independently of the voice model.
8 8
9## Understand the two parts9## Understand the two parts
10 10
11- **GPT-Live handles conversation.** It listens, speaks, and decides when to ask the backend for help. Give it a short prompt for conversation style and when to delegate.11- **GPT-Live handles conversation.** It listens, speaks, and decides when to ask the backend for help. Give it a short prompt for conversation style and when to delegate.
12- **The backend handles delegated tasks.** With Responses delegation, use a supported Responses model. With client delegation, connect any model, agent harness, or service your application runs. The backend reasons, uses tools, and returns results for GPT-Live to communicate. Keep detailed instructions, business rules, and tool workflows here.12- **The backend handles delegated tasks.** With Responses delegation, use a supported Responses model. With client delegation, connect any model, agent harness, or service your application runs. The backend reasons, uses tools, and returns results for GPT-Live to communicate. Keep detailed instructions, business rules, and tool workflows here.
13 13
14Your application owns permissions, confirmations, private function execution, and durable task state. Interrupting speech does not automatically cancel backend work. See [Voice agents](https://developers.openai.com/api/docs/guides/voice-agents) to compare GPT-Live with Realtime and chained voice applications.14Your application checks permissions, obtains required confirmations, runs functions that access your systems, and saves task progress. Backend work can continue when the caller interrupts the assistant; your application decides whether to finish or cancel it. See [Voice agents](https://developers.openai.com/api/docs/guides/voice-agents) to compare GPT-Live with Realtime and chained voice applications.
15 15
16 16
17 17
19 19
20## Choose how to run the backend20## Choose how to run the backend
21 21
22Start with **[Responses delegation](https://developers.openai.com/api/docs/guides/live-delegation?delegation-mode=responses#configure-responses-delegation)** when a managed backend fits: GPT-Live calls your configured Responses model, supplies conversation context, and returns results. Choose **[client delegation](https://developers.openai.com/api/docs/guides/live-delegation?delegation-mode=client#configure-client-delegation)** when your application needs to control backend execution, context, or which results reach GPT-Live.22Start with **[Responses delegation](https://developers.openai.com/api/docs/guides/live-delegation?delegation-mode=responses#configure-responses-delegation)** to have OpenAI run the backend model and pass conversation context and results between it and GPT-Live. Your application still runs your own function tools. Choose **[client delegation](https://developers.openai.com/api/docs/guides/live-delegation?delegation-mode=client#configure-client-delegation)** to connect an existing agent or control the backend’s context, execution, and returned results yourself.
23 23
24See [Choose a delegation mode](https://developers.openai.com/api/docs/guides/live-delegation#choose-a-delegation-mode) for the comparison and configuration details. Choose the mode when you create the session; to change modes, start a new session.24See [Choose a delegation mode](https://developers.openai.com/api/docs/guides/live-delegation#choose-a-delegation-mode) for the comparison and configuration details. Choose the mode when you create the session; to change modes, start a new session.
25 25
343. Wait for `session.started`, then speak and listen to a reply. Ask a question that needs current information to try the web search backend.343. Wait for `session.started`, then speak and listen to a reply. Ask a question that needs current information to try the web search backend.
354. End the conversation and [close the session](https://developers.openai.com/api/docs/guides/live-conversations#usage-and-graceful-close) to collect final usage and release the connection.354. End the conversation and [close the session](https://developers.openai.com/api/docs/guides/live-conversations#usage-and-graceful-close) to collect final usage and release the connection.
36 36
37Check both the spoken conversation and the backend result. A session-start event confirms startup; listening to a reply and checking the search result verify separate parts of the application.37For your first test, listen to the assistant and check that its answer reflects the backend’s search result.
38 38
39GPT-Live voice sessions are billed by duration, per second. See the [model pricing](https://developers.openai.com/api/docs/models/gpt-live-1) for the current rate. Backend model and tool usage is billed separately. See [Cost optimization](https://developers.openai.com/api/docs/guides/voice-latency-cost?api=live) for usage accounting and ways to reduce costs.39GPT-Live voice sessions are billed by duration, per second. See the [model pricing](https://developers.openai.com/api/docs/models/gpt-live-1) for the current rate. Backend model and tool usage is billed separately. See [Cost optimization](https://developers.openai.com/api/docs/guides/voice-latency-cost?api=live) for usage accounting and ways to reduce costs.
40 40
41## Choose a connection41## Choose a connection
42 42
43- **[WebRTC](https://developers.openai.com/api/docs/guides/voice-webrtc?api=live)** for browser voice applications. Media tracks carry audio; a data channel carries JSON events.43- **[WebRTC](https://developers.openai.com/api/docs/guides/voice-webrtc?api=live)** for browser voice applications. It carries microphone and speaker audio on media tracks and JSON events on a data channel.
44- **[WebSockets](https://developers.openai.com/api/docs/guides/voice-websockets?api=live)** for server-side audio integrations. The primary socket carries audio and control events.44- **[WebSockets](https://developers.openai.com/api/docs/guides/voice-websockets?api=live)** for server-side audio integrations. One connection carries audio and control events.
45- **[Server-side controls](https://developers.openai.com/api/docs/guides/voice-server-controls?api=live)** for backend access to an existing session. A sideband connection carries events while audio stays on the primary connection.45- **[Telephony and SIP](https://developers.openai.com/api/docs/guides/voice-sip?api=live)** for connecting phone calls.
46- **[Telephony and SIP](https://developers.openai.com/api/docs/guides/voice-sip?api=live)** for phone integration paths and provider guidance.46
47To monitor or control an existing session from your backend, add a [server-side connection](https://developers.openai.com/api/docs/guides/voice-server-controls?api=live). This additional WebSocket is called a **sideband**; audio continues through the session’s primary connection.
47 48
48## Partner integrations49## Partner integrations
49 50
50For applications built with **LiveKit**, **Twilio**, **Telnyx**, or **Daily/Pipecat**, start with the [partner integration overview](https://developers.openai.com/api/docs/guides/live-partner-integrations) to choose a connection for your existing media path.51If your application already uses **LiveKit**, **Twilio**, **Telnyx**, or **Daily/Pipecat**, follow the [partner integration overview](https://developers.openai.com/api/docs/guides/live-partner-integrations) to connect its existing calls or audio streams to GPT-Live.
51 52
52## Continue building53## Continue building
53 54