
发送长视频链接,或上传文件。它会读取源视频,自主判断值得剪辑的片段,生成带字幕的短片,审核结果,并只交付通过审核的短片——每条附有推荐标题、开场钩子和话题标签。

粘贴播客、访谈、直播或讲座视频的链接,或直接上传视频,即可拿到做好的竖屏短片:挑出能独立成立的片段,裁成 9:16,配逐词字幕,画面跟随说话人,音量调平,每条附标题、开场金句和发布文案。Google Drive 或 Dropbox 链接:直接交付 MP4 成片;YouTube 链接:先给出剪辑方案。

Pick a section of a video and get a high-quality GIF (palette-optimized), an animated WebP and a silent mp4 loop in one run. Set a target size such as KakaoTalk, Discord or Slack, and it lowers fps, resolution and color count step by step to find the best quality under that size. Meme captions (top/bottom), boomerang, speed change and square/vertical crop are applied in the same pipeline. Runs fully locally (uv + ffmpeg), no API key.

Paste your script. Get the one moment viewers leave, why, and a rewrite of only that moment — not a whole new video.

Upload a long podcast, lecture or interview and get up to 5 vertical 1080x1920 shorts back in the chat. ShortLoom picks segments by loudness, speech and motion, trims pauses, reframes to 9:16 by following faces, and burns in word-timed captions. Or name the moments yourself, like "3:10 to 3:50". Runs in the Capafy cloud, nothing to install. Measured: a 9:40 video became 5 captioned shorts in 62 s on 2 CPUs. It does not judge jokes or story beats.

Turn one long video plus its transcript into ready-to-post vertical shorts. Ranked by a published hook score, cut on sentence boundaries only, reframed to 1080x1920 with burned captions, and shipped with titles, captions and hashtags per clip. You also get cutplan.csv, publish.csv and the exact ffmpeg commands, so the same transcript and flags always produce the same plan. No speech recognition, no face tracking, no promises about reach.

Turn any long-form video transcript, podcast text, or topic idea into viral short-form video scripts, exact timestamp highlight maps, retention hooks, and CapCut/Premiere editing blueprints.

Turn one long video into 5 vertical shorts (from videos of 2.5+ minutes) with translated + original subtitles (EN, ZH, ES, KO, JA), plus titles, hooks and hashtags.
