GPT‑Live‑1 将 ChatGPT 自然、全双工的对话能力带入 API，让开发者能更好地控制语音智能体的说话与行动方式。
我们正式在 API 中推出 GPT‑Live‑1，为开发者提供一个强大、自然的语音模型，用于构建支持语音的应用和业务工作流。GPT‑Live‑1 最初在 ChatGPT 中亮相，它能够同时听和说，并且正如在 Codex 和 ChatGPT Work 中所见，它可以将更深层的推理和操作委派给它所搭配的模型和工具。
针对 GPT‑Live‑1 的 API 版本，我们重点打造了多项新能力，让开发者能够围绕自己的用户、工作流和目标来引导和定制语音体验。GPT‑Live‑1 的一项核心优势——流畅的中断处理——已经带来了实际业务影响：在早期评估中，Speak 发现 GPT‑Live‑1 让学习者在语言辅导老师回应之前有更多时间思考，与以往基于轮次的系统相比，中断减少了近 80%。
GPT‑Live‑1 在 API 中的关键优势：
- **中断处理：**通过单一模型同时对传入和传出音频进行推理，改进了中断处理，避免了链式 STT–LLM–TTS 架构的延迟和脆弱的交接。
- **推理与工具调用委派：**GPT‑Live‑1 可以将推理和工具调用委派给后端文本模型，例如 GPT‑6 Astra 或第三方模型。
- 语气、节奏和风格： 让开发者通过系统提示词来塑造智能体的语气、节奏和对话风格。
- 静默上下文管理与背景噪声： 更好地处理背景噪声和静默，不会打断对话，也不会把每一步都大声叙述出来。
- 长会话可靠性： 提升长时间交互中的上下文保持能力和对话质量。
- 电话支持： 支持部署用于电话通话的全双工语音智能体，从餐厅预订到客户支持均可覆盖。
Yelp Host 借助 GPT‑Live‑1 处理背景噪声、旁支对话和打断，顺畅地完成预订。
简化你的语音智能体架构，降低语音延迟
传统语音智能体将语音转文字、推理模型和文字转语音拼接在一起。每一次交接都会增加延迟，也带来更多丢失时机、上下文或对话自然节奏的可能。开发者往往要负责协调这些环节，包括当有人打断、停顿或改变话题时该如何处理。
GPT‑Live‑1 在单一模型中同时处理听与说，简化了语音层。它能在打断和确认应答发生时即时响应，同时将更深层的推理交给后端处理。这让对话在工作于后台进行的同时得以持续。
“与我们此前的级联式构建相比，GPT‑Live‑1 将我们的代码库简化了 80%，并移除了 23K 行代码。这实现了自然、实时的患者对话，并让我们的团队得以腾出手来改进从预约挂号到就医导航的体验。”
——Tony Stoyanov，联合创始人兼 CTO
开发者可选择对话背后的模型、工具和智能体框架。例如，他们可以将 GPT‑Live‑1 与 Luna 这样的模型搭配，用于排程或订单更新等高并发任务，并使用 Astra 这样的模型处理需要推理的复杂客户问题。这种灵活性让开发者能够为每项任务匹配相应的推理深度、速度和成本。
GPT‑Live‑1 原生提供 ASR 转写文本和响应文本。它还具备出色的字母数字理解能力，并支持关键词偏置。尽管 GPT‑Live‑1 并非基于轮次的模型，但它原生支持轮次检测，因此开发者仍可围绕明确的轮次边界进行构建。
衡量全双工优势
在我们的各项评估中，GPT‑Live‑1 相比 GPT‑Realtime‑2.1 将 Full Duplex Bench 的表现提升了 30 个百分点，在轮次转换延迟和交互行为方面均有大幅提升。搭配以中等推理强度运行的 GPT‑6 Astra，它还在 Tau3 上排名第一，该基准衡量的是前沿语音智能体在端到端任务上的智能水平。
评估航空、零售和电信领域的语音客服任务。Pass@1 衡量任务成功率；主指标对每个领域赋予相同权重。
- GPT Live 后端：Astra（medium）。
评估带有知识检索和账户工具的语音银行支持。Pass@1 是 97 项 banking_knowledge 任务中成功完成的比例。
- GPT Live 后端：Astra（medium）。
评估停顿处理、对话轮次转换、打断和反馈应答。
测试对背景语音、对他人说话、听者反馈应答以及打断的反应。
衡量智能体在用户结束一轮发言后多快开始回复。
测试从包含自然停顿、犹豫和自我纠正的语音请求中进行工具使用。Pass@1 对工具调用序列进行评分。
- GPT Live 后端：Terra（low）。
评估对使用工具请求的口头回答，其中包含停顿、犹豫和自我纠正。评分回答与参考意图的匹配程度。
- GPT Live 后端：Terra（low）。
客户评价
“将 GPT‑Live‑1 集成到 Yelp Host 和 Hatch 中，相比我们传统的语音架构，改善了轮次切换和准确性。当 Yelp Host 使用 GPT‑Live‑1 接听电话（如预订和订餐）时，我们看到通话处理率有了显著提升。来电者也在说出更完整、更自然的句子，这告诉我们电话另一端的体验确实有了本质的不同。”
—Alex Levy，首席技术官
“一位优秀的语言导师知道何时给学习者空间、何时介入，而 GPT‑Live‑1 将这种自然感带到了 Speak 的 Live Tutor Lessons 中——在我们的早期评估中，相比之前基于轮次的系统，它将思考停顿期间的中断减少了近 80%。”
—Andrew Hsu，联合创始人兼首席技术官
“GPT‑Live-1 展示了全双工模型能够释放的潜力：它将 AI 语音支持从走走停停的节奏推向电话通话般的自然流畅。客户可以自然地停顿、打断和改变方向；语音表现是明显的进步；Fin 可以将这种自然对话与其专有的支持系统相结合，完成解决问题所需的更深层工作。对我们来说，这是迄今为止关于语音支持走向的最清晰信号。”
——Jordan Neil，首席运营官
“有了 Devin 和 GPT‑Live‑1，与 AI 工程师共事开始更像是在和一位队友协作。你可以把想法聊透、对某个方案做压力测试，或者干脆在离开键盘时把工作交出去。”
——Walden Yan，联合创始人兼首席产品官
全新语音选项
开发者需要契合自身产品、并且对使用者来说听起来自然的语音。借助 GPT‑Live‑1，我们从小规模的实时语音集扩展到了覆盖不同口音、方言和语言的更广泛选择，让开发者在助手声音的呈现方式上拥有更多选择。
未来几个月，我们将继续扩展语音选项和语言可用范围。
定价与可用性
GPT‑Live‑1 即日起 在 API 中可用 ，前端语音层定价为每分钟 $0.05。将其与契合你产品的后端模型和智能体框架搭配，即可构建出能够随所需处理的工作量而扩展的语音体验。
将 GPT-Live-1 连接到 Codex
1import { Codex } from "@openai/codex-sdk";23const thread = new Codex().startThread({4 workingDirectory: "./repo",5 sandboxMode: "read-only",6 approvalPolicy: "never",7});89async function answer(live, delegationId, context) {10 const { finalResponse } = await thread.run(11 使用这个 repo 回答最新问题。12 用两句简短的口语作答。\n${context}13 );1415 live.send({16 type: "session.commentary.append",17 delegation_id: delegationId,18 content: finalResponse,19 });20}
将 GPT-Live-1 连接到 Codex。本摘录展示了应用程序如何将对话上下文传递给 Codex，并将其回答返回给 GPT-Live-1。连接设置与委派处理已省略。
如需自定义语音访问权限，请联系销售团队，了解有关资格条件和申请流程的更多信息。
在 GPT‑Live‑1 之上构建语音工作流的另一种方式是使用 OpenAI Presence，它利用该模型驱动实时语音交互。Presence 帮助企业在部署可信 AI 智能体时，使其能够回答问题、解决问题、使用公司系统、执行已批准的操作，并在需要时升级转交人工处理。如需了解更多信息，请联系您的 OpenAI 客户总监。
GPT‑Live‑1 brings ChatGPT’s natural, full-duplex conversations to the API, with more control over how voice agents speak and act.
We’re launching GPT‑Live‑1 in the API, giving developers a powerful, natural voice model for building voice-enabled apps and business workflows. First introduced in ChatGPT, GPT‑Live‑1 is capable of listening and speaking at the same time, and, as seen with Codex and ChatGPT Work , can delegate deeper reasoning and actions to the models and tools it is paired with.
For the API release of GPT‑Live‑1, we’ve focused on new capabilities that let developers steer and customize voice experiences around their users, workflows, and goals. A core GPT‑Live‑1 strength, smooth interruption handling, is already delivering business impact: in early evaluations, Speak found that GPT‑Live‑1 gave learners more time to think before the language tutor responded, cutting interruptions by almost 80% versus previous turn-based systems.
Key strengths of GPT‑Live‑1 in the API:
- **Interruption handling:**Improves interruption handling via a single model that reasons over incoming and outgoing audio together, avoiding the latency and brittle handoffs of chained STT–LLM–TTS architectures.
- **Reasoning & tool calling delegation:**GPT‑Live‑1 can delegate reasoning and tool calls to a backend text model like GPT‑6 Astra or a third-party model.
- Tone, pace, and style: Lets developers shape an agent’s tone, pace, and conversational style through the system prompt.
- Silent context management & background noise: Better handles background noise and silence without interrupting the conversation or narrating every step out loud.
- Long-session reliability: Improves context retention and conversational quality across extended interactions.
- Telephony support: Enables deployment of full-duplex voice agents for phone calls, from restaurant reservations to customer support.
Yelp Host seamlessly secures a reservation with GPT‑Live‑1 handling background noise, side conversations, and interruptions.
Simplify your voice-agent architecture and reduce voice latency
Traditional voice agents stitch together speech-to-text, a reasoning model, and text-to-speech. Each handoff adds latency and creates more opportunities to lose timing, context, or the natural rhythm of a conversation. Developers are often the ones left coordinating those stages, including what happens when someone interrupts, pauses, or changes direction.
GPT‑Live‑1 handles listening and speaking in a single model, simplifying the voice layer. It can respond to interruptions and acknowledgements as they happen, while delegating deeper reasoning to the back end. This lets the conversation continue while work happens in the background.
“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
Developers choose the models, tools, and agent harness behind the conversation. For example, they might pair GPT‑Live‑1 with a model like Luna for high-volume tasks like scheduling or order updates, and use a model like Astra for complex customer issues that require reasoning. That flexibility lets developers match reasoning depth, speed, and cost to each task.
GPT‑Live‑1 natively provides ASR transcripts and response text. It also offers strong alphanumeric understanding and supports keyword biasing. Although GPT‑Live‑1 is not a turn-based model, it natively supports turn detection, so developers can continue to build around explicit turn boundaries.
Measuring the full-duplex advantage
Across our evaluations, GPT‑Live‑1 improves Full Duplex Bench performance by 30 percentage points over GPT‑Realtime‑2.1, with large gains in turn-taking latency and interactive behavior. Paired with GPT‑6 Astra at medium reasoning effort, it also ranks #1 on Tau3, which measures frontier voice-agent intelligence on end-to-end tasks.
Evaluates spoken customer-service tasks in airline, retail, and telecom domains. Pass@1 measures task success; the headline gives each domain equal weight.
- GPT Live backend: Astra (medium).
Evaluates spoken banking support with knowledge retrieval and account tools. Pass@1 is the fraction of 97 banking_knowledge tasks completed successfully.
- GPT Live backend: Astra (medium).
Evaluates pause handling, conversational turn taking, interruptions, and backchannels.
Tests reactions to background speech, speech to another person, listener backchannels, and interruptions.
Measures how quickly the agent starts its reply after the user finishes a turn.
Tests tool use from spoken requests containing natural pauses, hesitations, and self-corrections. Pass@1 scores the tool-call sequence.
- GPT Live backend: Terra (low).
Evaluates the spoken answer to tool-using requests containing pauses, hesitations, and self-corrections. Scores how well the answer matches the reference intent.
- GPT Live backend: Terra (low).
What customers are saying
“Adding GPT‑Live‑1 into Yelp Host and Hatch improved turn-taking and accuracy over our traditional voice architecture. When Yelp Host uses GPT‑Live‑1 to answer calls, like reservations and food orders, we're seeing meaningful improvements in call handling rates. Callers are also speaking fuller, more natural sentences, which tells us the experience on the other end of the phone feels genuinely different.”
—Alex Levy, Chief Technology Officer
“A good language tutor knows when to give learners space and when to step in, and GPT‑Live‑1 brings that naturalness to Speak’s Live Tutor Lessons—in our early evaluations, it cut interruptions during thinking pauses by almost 80% compared with previous turn-based systems.”
—Andrew Hsu, Co-founder & CTO
“GPT‑Live-1 shows what a full-duplex model can unlock: It moves AI voice support from the stop-start rhythm toward the natural flow of a phone call. Customers can pause, interrupt, and change direction naturally; voice delivery is a clear step forward; and Fin can combine that natural conversation with its proprietary support system to do the deeper work needed to resolve the issue. For us, this is the clearest signal yet of where voice support is heading.”
—Jordan Neil, COO
“With Devin and GPT‑Live‑1, working with an AI engineer starts to feel more like collaborating with a teammate. You can talk through an idea, pressure-test an approach, or just hand off work while you’re away from your keyboard.”
—Walden Yan, Co-Founder and CPO
New voice options
Developers need voices that fit their product and sound natural to the people using it. With GPT‑Live‑1, we’re expanding from a small set of real-time voices to a broader selection across accents, dialects, and languages giving developers more choice in how their assistants sound.
We’ll continue to expand voice options and language availability over the coming months.
Pricing & Availability
GPT‑Live‑1 is available in the API today at $0.05 per minute for the front-end voice layer. Pair it with the backend model and agent harness that fit your product, then build a voice experience that can scale with the work it needs to do.
Connecting GPT-Live-1 to Codex
1import { Codex } from "@openai/codex-sdk";23const thread = new Codex().startThread({4 workingDirectory: "./repo",5 sandboxMode: "read-only",6 approvalPolicy: "never",7});89async function answer(live, delegationId, context) {10 const { finalResponse } = await thread.run(11 Answer the latest question using this repo.12 Reply in two short spoken sentences.\n${context}13 );1415 live.send({16 type: "session.commentary.append",17 delegation_id: delegationId,18 content: finalResponse,19 });20}
Connecting GPT-Live-1 to Codex. This excerpt shows how an application passes conversation context to Codex and returns its answer to GPT-Live-1. Connection setup and delegation handling are omitted.
For custom voice access, contact sales to learn more about eligibility and the request process.
Another way to build voice workflows on top of GPT‑Live‑1 is with OpenAI Presence, which uses the model to power real-time voice interactions. Presence helps enterprises deploy trusted AI agents that can answer questions, resolve issues, use company systems, take approved actions, and escalate to people when needed. Reach out to your OpenAI account director to learn more.