GPT-Live-1 的声音现在可以由你描述：提示词转语音功能上线，九种国际预设同步登场

For most of the year, choosing a voice for G​PT-Live-1 meant picking a string. You read a list, you set audio.output.voice/ to one of them at session start, and if none of them fit the brand you were building, you shipped with the closest one. That changed on 16 October 2026, when O​penAI released prompt-to-voice to all API customers and added nine natural voices aimed at five non-English markets. The new capability is small on paper — one endpoint, free to call — and it rewrites the earliest decision most voice products make. It also lands a day after G​PT-Live-1.1 picked up native Decisions integration inside the same Live sessions, which matters because the two changes push in opposite directions: one makes the voice more bespoke, the other makes the thing behind the voice more active.

这是实际发布的内容、它的成本，以及新语音列表中两处任何发布说明都未提及的不对称之处。

10 月 16 日发布版包含的内容

有两件事，都标记到 gpt-live-1/ 模型以及两个端点——v1/audio/voices/ 和 v1/live/sessions/。

• Prompt-to-voice — create a custom voice from a text description, through the API or the voice Playground, then reuse it in G​PT-Live sessions. Released for all API customers.

• 九种全新的自然语音——Nuri、Juni 和 Flitz（德语）；Brise 和 Sillage（法语）；Nira（印地语）；Harema 和 Shida（日语）；Noeul（韩语）。

• Cost — creating a prompt-generated voice is free. Using it speech-synthesises at regular G​PT-Live session rates, and auditioning it in the Playground counts as a normal session for billing.

• A day earlier, on 15 October — native Decisions integration in G​PT-Live, so an app can act while the user is still speaking or react to what Live says, via native delegation or transcript deltas.

The pricing context that makes those numbers legible: a G​PT-Live-1 voice session runs $0.05 per minute, billed by the second and not rounded up. That line is unchanged by this release, and no part of prompt-to-voice changes it.

API 调用就是一次 POST，而它的形态能告诉你模型在做什么

你将 JSON 发送到/v1/audio/voices，并附带type: "prompt"、name、prompt 描述，以及 "model": "auto"。可选的 script_hint（最多 2,000 个字符）可提供示例语音来引导生成；如果省略它，服务会自行编写脚本，但空白或非常短的脚本会被拒绝。

该响应不包含音频。它只返回元数据——对象类型、id、name、created_at（Unix 秒）、type，以及你的原始提示。这是一个有意为之的设计选择，并带来一个实际后果：创建时没有任何可预览的内容，因此你只能像实际使用它那样试听，而这会按一次会话计费。每次请求都会保存一个全新的语音，所以请将 ID 和描述放在一起——描述是你所请求内容的唯一记录。

The controllable qualities are all expressed as ordinary language rather than as separate parameters: accent, age, presentation, pitch, timbre, texture, loudness, speech rate and distinctiveness. O​penAI's own examples include an accent phrased as "American English, with a strong accent" and a pace phrased as "Medium pace." There is no numeric prosody slider anywhere in the surface, which means the prompt is the API.

这是一个实实在在的权衡。滑块是可复现的；句子则不是。两个团队写同一段描述，可能无法获得相同的语气，而且文档明确指出，生成的语气并不保证独一无二。就像你把系统提示词纳入版本控制一样，也要把提示词文本纳入版本控制。

值得围绕其进行设计的限制

• 每个组织 100 个由提示词生成的语音。该上限也适用于你的客户，因此，如果你为每个最终用户生成一个语音，100 就是你整个租户群的上限，而不是每账户的配额。

• G​PT-Live only. Prompt-generated voices do not work with the Text to Speech, Realtime, or Chat Completions APIs. A voice you describe today is a voice you cannot reuse in a batch narration pipeline tomorrow.

• 不可与零数据保留同时使用。通过 /v1/audio/voices创建语音（包括录制路径）不在支持范围内。

• 仅限成人，且不得模仿。不允许使用儿童声音，指导原则是描述声音特质，而非提及真实人物或角色的名字。

• 基于录音的声音走的是另一条更严格的路径——每个组织最多 20 个，样本时长不超过 30 秒，格式为 mpeg、wav、ogg、aac、flac、webm 或 mp4，至少包含 5 秒真实语音和 15 个转录 token，上传大小上限为 10 MiB，并且需要提供同意录音。

The 100-voice cap and the G​PT-Live-only restriction are the two that shape architecture. If you are a platform generating a voice per customer, you have a hard ceiling and no fallback path into the cheaper speech APIs.

完整的预设列表，以及它在哪些方面较为薄弱

The nine additions land on top of a library that O​penAI's Live voice guide now lists twenty-one deep, plus marin/ as the default. Read as a whole, the list says something about who O​penAI expects to be building:

• 英语——六种区域色彩，而非一种。Quartz 和 Ripple 是澳大利亚的；Vesper 是英国的；Willow 和 Stone 是爱尔兰的；Gleam 和 Meridian 是北美的；Beacon 是菲律宾的；Delta 和 Cinder 是美国南部的。其中四种被标记为“生成”而非“自然”。

• 葡萄牙语 — Bossa 和 Tempo，均为巴西葡语，均为 Natural。

• 德语——Nuri、Juni 和 Flitz。两个阴性，一个阳性。

• 法国 —— Brise 和 Sillage。一款女香，一款男香。

• 日语——Harema 和 Shida。一个偏女性化，一个偏男性化。

• 印地语——Nira。仅女性。

• 韩语 — Noeul。仅限女性。

Two asymmetries fall out of that. First, German gets three voices while French, Japanese, Hindi and Korean get two, two, one and one — an allocation that tracks O​penAI's European market priorities more than it tracks speaker populations. Second, Hindi and Korean each get exactly one voice, and both are feminine. If your product needs a masculine Hindi or Korean voice from the preset library, there isn't one, and your options are a recording-based voice or a prompt-generated one. That is precisely the gap prompt-to-voice was released to fill, arriving the same day — which suggests the sequencing is not a coincidence.

指南还指出了一个容易被误读的地方：“区域影响”描述的是某个声音的说话风格，而不是对它能够说哪些语言的硬性限制。应当针对你的应用实际遇到的语言和发音来测试候选声音，而不是想当然地认为带德语标签的声音就是德语输入的合适选择。

你无法在通话中途更改的设置。语音是启动时设定。你要在创建会话时指定 audio.output.voice/，而要更改它，就得启动一个新会话。分叉会继承源会话的模型、原始指令和输入，因此比较两种语音意味着基于同一配置创建两个会话，而不是在一个会话内切换。如果你要针对真实用户对语音进行 A/B 测试，那就意味着需要两个会话工厂，而不是一个带开关的会话——这也意味着每一次比较都会在两条分支上耗费完整的会话时间。

后端位于何处，以及为什么这是该决策中成本较低的那一半

The voice is the half of a Live session that users hear. The other half is delegation: G​PT-Live listens and speaks while a backend model or agent does the reasoning and tool work, and backend usage is billed separately at normal rates. O​penAI's own cost guidance makes the arithmetic blunt — full active session time is billable including while the backend works, and using a faster backend model with parallel tool calls so the user finishes a minute earlier saves $0.05 in voice charges. The voice rate is fixed per minute; the backend is the dial.

That is where a router earns its place rather than being bolted on. Client delegation lets you point a G​PT-Live session at your own agent, harness or service — which means the delegated backend can be any model, and the useful version of "any model" is one key and one endpoint rather than a contract per provider. OrcaRouter fronts 200-plus models at provider list price with no markup, so a backend price cut reaches you the same day it lands, and automatic failover covers a provider having a bad afternoon without the user hearing it. The routing DSL lets you compose the reasoning step out of several models instead of betting the conversation on one, which is the shape you want for a delegated backend that has to answer inside a live conversation's latency budget.

One boundary to state plainly: OrcaRouter does not host G​PT-Live-1, and nothing in this release changes that. The router's role in a Live architecture is behind the delegation boundary, not in front of the voice.

本周做什么

如果你已经在预设语音上发布了一款 Live 产品，那这九项新增内容就是一张免费通行证，让你得以测试自己的非英语用户是否真的得到了服务——在一小部分会话中更换语音，然后倾听，同时要记住，每次更换都会为每个实验组额外消耗一次会话。

如果你曾因品牌契合度而受阻，那么 prompt-to-voice 现在值得花一天去试验，但有一个日后会让你吃苦头的注意事项：把描述文本当作构建产物。给它做版本管理，把返回的 ID 放在它旁边，并假定你将来需要重新生成——因为在组织上限为 100 个音色、且创建时没有预览的情况下，你周二喜欢的那个音色，是你必须能够有意复现出来的东西。

And if you are building the whole system rather than just the voice, the more consequential October change may be the one nobody wrote a blog post about. A day before the voices landed, G​PT-Live picked up native Decisions integration — the ability to act while the user is still mid-sentence. A voice that fits is a product decision. A backend that answers before the user finishes asking is an architecture one.

这一切都并非假设：像本次这样的发布背后那些有据可查的 AI 故障与智能体事件，均收录于 OrcaRouter AI 事件档案库中，其中每条记录都附有自己的来源、严重程度、置信度评级和存疑标记。