ElevenLabs 在观云中负责什么
ElevenLabs 是观云当前接入的语音服务，负责转写和配音相关工作流。Scribe v2 用于视频、音频文件和符合条件的源链接语音转文字；字幕配音链路可以使用配置好的 ElevenLabs 声音把翻译后的字幕合成音频。这个页面会把语音模型与 DeepSeek 文本规划、生图生视频模型分开说明。
- 模型
- ElevenLabs Scribe v2
- 接入方
- ElevenLabs
- 模型 ID
- scribe_v2
- 官方资料
- ElevenLabs
模型能力
Speech-to-text transcription for uploaded and source-url media using ElevenLabs Scribe v2.
Word-level timing and segment normalization for subtitles, transcripts, and summary timelines.
Optional diarization support when the deployment enables speaker-aware transcription.
Text-to-speech dubbing support through ElevenLabs voices and the configured ElevenLabs TTS model.
适合的真实任务
Video summary and subtitle jobs that need accurate transcripts before DeepSeek creates the report.
Captioning workflows where timestamps must stay aligned with the source media.
Subtitle dubbing tasks that synthesize translated lines back into voice audio.
推荐工作流
- 01
Submit a video, audio file, or supported source URL to the summary or subtitle pipeline.
- 02
Use ElevenLabs Scribe v2 for transcription and normalize words, speakers, and segment timestamps.
- 03
Pass the transcript into downstream subtitle rendering, summary generation, translation, or chat flows.
- 04
For dubbing, choose an ElevenLabs voice and synthesize translated subtitle lines into output audio.
ElevenLabs 工作流示例
创作者上传一场 webinar，观云把音频交给 ElevenLabs Scribe v2 转写，再把时间轴归一化成字幕，同时把转写文本交给 DeepSeek 生成总结，并可继续生成翻译配音。
边界和审核
ElevenLabs is used for speech transcription and dubbing, not for text reasoning or media planning.
Large files, long durations, diarization, and source-url transcription depend on provider limits and deployment configuration.
相关入口
FAQ
观云是否已经接入 ElevenLabs？
是。观云使用 ElevenLabs 进行语音转文字；当部署配置完成时，也可以使用 ElevenLabs 声音做字幕配音。
ElevenLabs 和 DeepSeek 是同一种模型吗？
不是。ElevenLabs 负责语音转写和语音合成；DeepSeek 负责文本推理、总结、文案和规划。