https://channel.nostses.icuhttps://channel.nostses.icu/posts/1994https://channel.nostses.icu/posts/1994Fri, 09 Jan 2026 08:12:36 GMT<div class="tgme_widget_message_forwarded_from accent_color">Forwarded from <a class="tgme_widget_message_forwarded_from_name" href="https://t.me/iokhk/8368" target="_blank" rel="noopener"><span>OKHK <i class="emoji" style="background-image:url('//telegram.org/img/emoji/40/F09F9180.png')"><b>👀</b></i></span></a></div><b><u>Vibe </u></b><mark class="highlight"><b><u>Voice</u></b></mark><b><u> - 微软开源 AI 语音框架</u></b><br /><br /><a href="https://github.com/microsoft/VibeVoice" target="_blank" rel="noopener" title="https://github.com/microsoft/VibeVoice">https://github.com/microsoft/VibeVoice</a><br /><br />VibeVoice 是微软开源的 AI 语音框架，用于从文本生成富有表现力的长篇多说话人对话音频（如播客），通过 LLM 理解文本上下文和对话流，生成高保真音频。<br /><br /><b>● </b>长篇多说话人模型：支持最长 90 分钟音频、最多 4 个不同说话人的对话式语音合成<br /><b>● </b>实时流式 TTS 模型：初始音频延迟约 300 毫秒，支持流式文本输入的实时语音生成<br /><b>● </b>超低帧率语音分词器：以 7.5Hz 帧率的连续语音分词器保持音频保真度同时提升计算效率<br /><b>● </b>多语言支持：支持英文、中文以及实验性支持 9 种语言（德、法、意、日、韩、荷、波、葡、西班牙语）<br /><b>● </b>风格化语音：提供 11 种不同的英文风格语音供用户选择<br /><br /><a href="/search/result?q=%23TTS" title="#TTS">#TTS</a> <a href="/search/result?q=%23Voice" title="#Voice">#Voice</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a><a class="tgme_widget_message_link_preview" href="https://github.com/microsoft/VibeVoice" target="_blank" rel="noopener" title="Open-Source Frontier Voice AI. Contribute to microsoft/VibeVoice development by creating an account on GitHub."> <div class="link_preview_site_name accent_color">GitHub</div> <img class="link_preview_image" alt="GitHub - microsoft/VibeVoice: Open-Source Frontier Voice AI" src="/static/https://cdn4.telesco.pe/file/DzS8oWz_onmUzlaZ3UagsV6IwOyoPS1caq8HI3ItDpYglSnYe6Vo26_jpAgdmnnsKOOJbnit77p0H8CrSBPfP71Fxe2dYN_Iz3Rwp7-XhNk1uvNU3Pj4ySt7HWwpG3x9GwJAQk2HXY13Cnsslac9G_zgEhTCqGJgeLCimIXxs8Na5GFtXJIcB_BGepA3wimfnOe5Pff3Lag7xGpLXd4dbotjkQb141fIFUWEJ6DOJEnkBs82aRgDQnvjf-3sFaF9xNiFwjS7UjQ558IgGnpCsNnn64enQB3MrVtPyuG0pgu6vHgm9q2aUK_toZjRAa1vvFL2PPuVhmTjhl8RtnIpqA.jpg" width="1200" height="630" loading="eager" /> <div class="link_preview_title">GitHub - microsoft/VibeVoice: Open-Source Frontier <mark class="highlight">Voice</mark> AI</div> <div class="link_preview_description">Open-Source Frontier <mark class="highlight">Voice</mark> AI. Contribute to microsoft/VibeVoice development by creating an account on GitHub.</div> </a>https://channel.nostses.icu/posts/1708https://channel.nostses.icu/posts/1708Sun, 26 Oct 2025 12:48:48 GMT<div class="tgme_widget_message_forwarded_from accent_color">Forwarded from <a class="tgme_widget_message_forwarded_from_name" href="https://t.me/iokhk/7442" target="_blank" rel="noopener"><span>OKHK <i class="emoji" style="background-image:url('//telegram.org/img/emoji/40/F09F9180.png')"><b>👀</b></i></span></a></div><b><u>蛐蛐 (QuQu) - 开源免费的 Wispr Flow 替代方案 | 为中文而生的下一代智能语音工作流</u></b><br /><br /><a href="https://github.com/yan5xu/ququ" target="_blank" rel="noopener" title="https://github.com/yan5xu/ququ">https://github.com/yan5xu/ququ</a><br /><br /><blockquote>离线语音输入，无需会员，不连云端，保证隐私。</blockquote><br /><br />蛐蛐 (QuQu) 是 一个开源免费的 Wispr Flow 替代方案, 专为中文用户打造的注重隐私的桌面端语音输入与文本处理工具。<br /><br />它集成了阿里巴巴开源的 FunASR 本地语音识别模型, 并支持配置国内顶尖的大语言模型, 提供高精度的中文语音转文字和智能文本优化功能。<br /><br /><b>● </b>完全免费开源, 数据本地处理, 保护用户隐私<br /><b>● </b>专为中文优化, 支持国产 AI 模型如通义千问、Kimi 等<br /><b>● </b>独创 "ASR 精准识别 + LLM 智能优化 " 的两段式引擎<br /><b>● </b>能准确识别并格式化编程术语, 支持自定义 AI 指令<br /><b>● </b>提供全局快捷键唤醒, 实时识别和无缝粘贴功能<br /><br /><a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23Voice" title="#Voice">#Voice</a> <a href="/search/result?q=%23Tool" title="#Tool">#Tool</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a><a class="tgme_widget_message_link_preview" href="https://github.com/yan5xu/ququ" target="_blank" rel="noopener" title="开源免费的 Wispr Flow 替代方案 | 集成FunASR本地模型和可配置大语言模型的下一代中文桌面语音工作流 - yan5xu/ququ"> <div class="link_preview_site_name accent_color">GitHub</div> <img class="link_preview_image" alt="GitHub - yan5xu/ququ: 开源免费的 Wispr Flow 替代方案 | 集成FunASR本地模型和可配置大语言模型的下一代中文桌面语音工作流" src="/static/https://cdn4.telesco.pe/file/eht6TGW-QWUMq6DjewlIRvf3tmnhMT1MDwCZLnHJMlKkUTb5t83XRSrUoTTBS09kX5pqPprogeoERG3hTwJKYU5Sx1ll1GN71fB1qyGrcNKZRNfgRHnLxJW25VxxL-hta4XzoR0LJBn-779Ey_sSbtXksCFcevuAh6PdAJMFrbTo701O9zMdyitvc3CpL_9M0_uaha2A2kZQaHyPTcv9kXltcGfl3zKqUNMk-nlE25TrZkE2P7U2s7LZONEgQcw8TtOwMDXxWz8Q3n-igIe5dw2dNunh5iWF3K0aaNAF91hB1ycKmAhfnktJNc1ZgQN4Yvw7007xqb5My5EKB4JeiQ.jpg" width="1200" height="630" loading="eager" /> <div class="link_preview_title">GitHub - yan5xu/ququ: 开源免费的 Wispr Flow 替代方案 | 集成FunASR本地模型和可配置大语言模型的下一代中文桌面语音工作流</div> <div class="link_preview_description">开源免费的 Wispr Flow 替代方案 | 集成FunASR本地模型和可配置大语言模型的下一代中文桌面语音工作流 - yan5xu/ququ</div> </a>https://channel.nostses.icu/posts/1704https://channel.nostses.icu/posts/1704Sun, 26 Oct 2025 12:43:03 GMT<div class="tgme_widget_message_forwarded_from accent_color">Forwarded from <a class="tgme_widget_message_forwarded_from_name" href="https://t.me/iokhk/7317" target="_blank" rel="noopener"><span>OKHK <i class="emoji" style="background-image:url('//telegram.org/img/emoji/40/F09F9180.png')"><b>👀</b></i></span></a></div><b><u>WhisperLiveKit - 本地语音转录工具（开源语音识别转文字）</u></b><br /><br /><a href="https://github.com/QuentinFuxa/WhisperLiveKit" target="_blank" rel="noopener" title="https://github.com/QuentinFuxa/WhisperLiveKit">https://github.com/QuentinFuxa/WhisperLiveKit</a><br /><br />WhisperLiveKit 是一个开源的本地语音转录工具，集实时语音转文本、翻译和说话人分离于一体。 它使用了最新的语音识别和语音分离技术，可以提供低延迟和高准确度的转录结果，能用于会议实时记录和跨语言交流。<br /><br /><b>● </b>提供实时语音转文本功能，支持多达 100 种语言的翻译<br /><b>● </b>支持实时说话人识别和分离<br /><b>● </b>提供简单易用的前端和后端 API<br /><b>● </b>支持在浏览器和 Chrome 扩展中使用<br /><b>● </b>支持在 Docker 容器中部署<br /><br /><a href="/search/result?q=%23Voice" title="#Voice">#Voice</a> <a href="/search/result?q=%23AI" title="#AI">#AI</a> <a href="/search/result?q=%23Tool" title="#Tool">#Tool</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a><a class="tgme_widget_message_link_preview" href="https://github.com/QuentinFuxa/WhisperLiveKit" target="_blank" rel="noopener" title="Simultaneous speech-to-text models. Contribute to QuentinFuxa/WhisperLiveKit development by creating an account on GitHub."> <div class="link_preview_site_name accent_color">GitHub</div> <img class="link_preview_image" alt="GitHub - QuentinFuxa/WhisperLiveKit: Simultaneous speech-to-text models" src="/static/https://cdn4.telesco.pe/file/kZZqy5JKyQjmEiaMnm8tot2adoLD6ZT93DY38npwuP4a6jDdw9dIGEnZBpgcPhBc8HGlc3YrJ34AgwE1FJHfqkG5LDRxQ66lciy8AgtFtGjzEJ5j563rlRUfuzW-fMnGJnTprPi6J2h5dISlyFSc-gspgOklXBLOfN9RsOJf9uJbNb9eU07IZqS9BnfBW3jjvtJY8Eu59TFTZHS_gPZdh_-SnCZTAv2-vlcgxiLcmq7FSQRdFBt3FC2yigXL5kGTnwT0Wdfg3y1Yd0PtBr8lSZyzEO4ENeiy5oxn1ihCuxmsoV0JjYGwxnQ-MylQGKoD1He1Un-sFkco2Im9JJHKKQ.jpg" width="1200" height="630" loading="eager" /> <div class="link_preview_title">GitHub - QuentinFuxa/WhisperLiveKit: Simultaneous speech-to-text models</div> <div class="link_preview_description">Simultaneous speech-to-text models. Contribute to QuentinFuxa/WhisperLiveKit development by creating an account on GitHub.</div> </a>https://channel.nostses.icu/posts/1403https://channel.nostses.icu/posts/1403Thu, 21 Aug 2025 06:18:16 GMT<div class="tgme_widget_message_forwarded_from accent_color">Forwarded from <a class="tgme_widget_message_forwarded_from_name" href="https://t.me/iokhk/6581" target="_blank" rel="noopener"><span>OKHK <i class="emoji" style="background-image:url('//telegram.org/img/emoji/40/F09F9180.png')"><b>👀</b></i></span></a></div><b><u>Klic Studio - 音视频处理工具（翻译 / 配音 ）</u></b><br /><br /><a href="https://github.com/krillinai/KlicStudio" target="_blank" rel="noopener" title="https://github.com/krillinai/KlicStudio">https://github.com/krillinai/KlicStudio</a><br /><br /><i>多平台发布 / 搬运视频好帮手</i><br /><br />Klic Studio 是由 Krillin AI 开发的多功能音视频本地化和增强解决方案。<br /><br />集成了视频翻译、配音和语音克隆，支持横屏和竖屏格式，确保在所有主要平台（Bilibili、小红书、抖音、微信视频、快手、YouTube、TikTok 等）上完美呈现。<br /><br />通过端到端的工作流程，您只需几次点击即可将原材料转化为精美的跨平台内容。<br /><br /><b>● </b>视频获取：支持 yt-dlp 下载或本地文件上传<br /><b>● </b>准确识别：基于 Whisper 的高精度语音识别<br /><b>● </b>智能分段：使用 LLM 进行字幕分段和对齐<br /><b>● </b>术语替换：一键替换专业词汇<br /><b>● </b>专业翻译：基于上下文的 LLM 翻译，保持自然语义<br /><b>● </b>语音克隆：提供 CosyVoice 中选择的语音音调或自定义语音克隆<br /><b>● </b>视频合成：自动处理横屏和竖屏视频及字幕布局<br /><b>● </b>跨平台：支持 Windows、Linux、macOS，提供桌面和服务器版本<br /><br /><a href="/search/result?q=%23Video" title="#Video">#Video</a> <a href="/search/result?q=%23Voice" title="#Voice">#Voice</a> <a href="/search/result?q=%23Tool" title="#Tool">#Tool</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a><a class="tgme_widget_message_link_preview" href="https://github.com/krillinai/KlicStudio" target="_blank" rel="noopener" title="AI video translation &amp; dubbing tool for humans and AI Agents, powered by LLMs. Full pipeline: download, transcribe, translate, TTS dub, reformat, cover generation. 100+ languages, optimized..."> <div class="link_preview_site_name accent_color">GitHub</div> <img class="link_preview_image" alt="GitHub - krillinai/KrillinAI: AI video translation &amp; dubbing tool for humans and AI Agents, powered by LLMs. Full pipeline: download…" src="/static/https://cdn4.telesco.pe/file/iO3RU-FznoI6ZtVg_D55v4B_x1DoQ8v9JMzWRf9fDb1wQZqhJaBO69JU0hjD8GZyUeJvd3RFkwnzBwLWKSY-wYPnxdU0OCrmGggtmssHKXBcDaV9oi_bMglB9RgLMDgFLzPGw-iTtEGLfACm3XT13-6-1IdBkchBua5JFf1dGctEvcsqp-SmuP7UpY3Qn_Vyb58_J3O7dxMJOYz47KVb68Ue7CiPD05O-vW-cgB61hp4ecj1j0O_9ueIiAdE_AtZIcVlyjTtaWtzgQqQpQoQ--o0JP4xRFeL8vxHQt_rpBy-_AjZ7DmXaPDbR2zW4WQ2iFnrAed-6IjA_vFeKiMogg.jpg" width="1200" height="630" loading="eager" /> <div class="link_preview_title">GitHub - krillinai/KrillinAI: AI video translation &amp; dubbing tool for humans and AI Agents, powered by LLMs. Full pipeline: download…</div> <div class="link_preview_description">AI video translation &amp; dubbing tool for humans and AI Agents, powered by LLMs. Full pipeline: download, transcribe, translate, TTS dub, reformat, cover generation. 100+ languages, optimized...</div> </a>https://channel.nostses.icu/posts/1391https://channel.nostses.icu/posts/1391Tue, 19 Aug 2025 18:12:26 GMT<div class="tgme_widget_message_forwarded_from accent_color">Forwarded from <a class="tgme_widget_message_forwarded_from_name" href="https://t.me/iokhk/6577" target="_blank" rel="noopener"><span>OKHK <i class="emoji" style="background-image:url('//telegram.org/img/emoji/40/F09F9180.png')"><b>👀</b></i></span></a></div><a href="https://github.com/wangwangit/tts" target="_blank" rel="noopener" title="https://github.com/wangwangit/tts">https://github.com/wangwangit/tts</a><br /><br />一个基于 Microsoft Edge TTS 的免费在线语音合成服务，可自行部署在 Cloudflare Worker 上<br /><br /><a href="/search/result?q=%23TTS" title="#TTS">#TTS</a> <a href="/search/result?q=%23Voice" title="#Voice">#Voice</a> <a href="/search/result?q=%23GitHub" title="#GitHub">#GitHub</a> <a href="/search/result?q=%23Cloudflare" title="#Cloudflare">#Cloudflare</a><a class="tgme_widget_message_link_preview" href="https://github.com/wangwangit/tts" target="_blank" rel="noopener" title="一个基于 Microsoft Edge TTS 的免费在线语音合成服务，支持 20+ 种中文声音，一键将文字转换为自然流畅的语音。 - wangwangit/tts"> <div class="link_preview_site_name accent_color">GitHub</div> <img class="link_preview_image" alt="GitHub - wangwangit/tts: 一个基于 Microsoft Edge TTS 的免费在线语音合成服务，支持 20+ 种中文声音，一键将文字转换为自然流畅的语音。" src="/static/https://cdn4.telesco.pe/file/Q32-8LqLlKH-2hY-qhC5qCL-C0xBXx1A0QgbKj5DetBsfyITNlAsl77caVE_r5Yt4v06wED2stVSckpGOFKSwBGjgSjhGHO-XfIlRSeJrs5HLwQEj9jKCGKmW7tAxwVcF3ue8LCzZxFL64Hzf3igzD6qm1n4xQCHp9AyJEiGCTBjoTd1qthP_FlGe4i2JTztLmLFxCQT2sYst1SBLdeXh6cEB5d43eXLd6t28aEAqLQ8FpaUWa-UjldYYQvU0M1HYgYX7W5MYxS987pFvUWVJhPW0MbEEXdxsQpo-dxPq9mQTfmaR6T8unM-aYgAU6uQrvU8TyZO1zHNMZWgybTbfA.jpg" width="1200" height="630" loading="eager" /> <div class="link_preview_title">GitHub - wangwangit/tts: 一个基于 Microsoft Edge TTS 的免费在线语音合成服务，支持 20+ 种中文声音，一键将文字转换为自然流畅的语音。</div> <div class="link_preview_description">一个基于 Microsoft Edge TTS 的免费在线语音合成服务，支持 20+ 种中文声音，一键将文字转换为自然流畅的语音。 - wangwangit/tts</div> </a>