{"schema":"hottub-news/story/v1","story":"st_7j5ew3mismylsss7jogq","title":"ElevenLabs 推出 v4 和 v4 Turbo 语音模型：支持 90 余种语言，10 秒素材即可克隆声音","title_en":"ElevenLabs releases v4 and v4 Turbo voice models: supports over 90 languages, clones voice from 10-second clip","outlets":1,"brief":null,"page":"https://hottub.news/stories/elevenlabs-releases-v4-and-v4-turbo-voice-models-supports-over-90-7j5ew3mism","count":1,"items":[{"id":"n_7j5ew3mismylsss7jogq","seq":546283,"kind":"article","title":"ElevenLabs 推出 v4 和 v4 Turbo 语音模型：支持 90 余种语言，10 秒素材即可克隆声音","title_en":"ElevenLabs releases v4 and v4 Turbo voice models: supports over 90 languages, clones voice from 10-second clip","url":"https://www.ithome.com/1/008/080.htm","summary":"IT之家 9 月 29 日消息，ElevenLabs 于当地时间周一正式推出了两款全新语音模型，分别为 ElevenLabs v4 与 v4 Turbo。新版模型强化了情绪表达控制能力，降低了语音智能体的响应延迟，同时支持 90 余种语言。 该公司去年发布了 v3 模型，并于今年早些时候在华沙举办的活动上对新一代模型进行了预热。v4 系列采用全新架构，实现了更精细的语音控制以及速度更快的声音克隆。官方表示，借助 v4，用户仅需 10 秒音频素材即可完成语音克隆。 在内容创作方面，当朗读大段文本时，新版模型可以更好地维持说话人的音色特征；朗读过程中还能够记住前后文本语境，随之调整语气。ElevenLabs 曾在 v3 版本当中引入内联标签用以设定语音情绪；到了 v4，标签功能得到进一步扩充，用户可以叠加多个标签，模型将按照标签的先后顺序执行。 旧版模型支持 70 种语言。经过优化之后，新版本的语言支持数量提升至 90 种。这家初创企业称，日语、巴西葡萄牙语、普通话以及粤语的合成质量提升最为显著。 过去一年，ElevenLabs…","published":"2026-09-28T23:29:37Z","seen":"2026-09-29T00:05:23Z","lang":"zh","country":"CN","source":{"id":"kite-ithome-com-f1cdd0","name":"ithome.com","domain":"ithome.com","outlet":{"wikidata":"Q28413779","name":"IThome","type":"website","country":"CN","designations":["website"]}},"via":"kite-ithome-com-f1cdd0","entities":[{"id":"Q116798355","name":"ElevenLabs","type":"org"}],"story":"st_7j5ew3mismylsss7jogq","category":"science-technology","category_p":0.949999988079071,"sentiment":"positive","tone":0.3799999952316284,"political":0.019999999552965164}]}