TonySkill

Viral clips

An AI finds the best moments of a long video (podcast, interview, talk, up to 45 minutes) and prepares them as 9:16 clips, with subject tracking and animated captions. Free, no watermark. Your video's sound is sent to TonySkill's server to be processed, then deleted.

Tool in improvement

Leave your email and we'll let you know as soon as it's ready.

Your address is only used to tell you when this tool is back.

How to turn a long video into clips

  1. Drop your long video (up to 2 GB and 45 minutes) or paste a YouTube link: the sound is sent to our server to be transcribed.

  2. Pick the number and length of clips, then “Find the best moments”: the AI suggests moments with a title and a caption.

  3. For each moment, “Create the vertical clip”: it opens in the editor in 9:16; add subject tracking and animated captions, then export.

Why make your clips here

Moments picked by AI
An AI reads the whole transcript and spots moments that stand on their own and hook from the first second, with a title and a caption.
Vertical clip in two clicks
Each moment opens in the editor, trimmed and in 9:16, with subject tracking and animated captions one click away.
No watermark, no subscription
As many clips as you like, no credits, no logo added.
Deleted after processing
Only the sound of your video goes to our server to be transcribed, then it's deleted. The video is only sent if you ask for the MP4 clips, and is deleted within 45 minutes at most. Never shared.

Something wrong with this tool?

Frequently asked questions

How do I make Shorts from a long video?

Drop the video (or paste a YouTube link): our server transcribes the speech, then an AI reads the text and picks moments that stand on their own and hook fast. You get the clips as MP4 files, or open each moment in the editor, already trimmed to 9:16, to add subject tracking and animated captions in one click each.

Is my video sent to a server?

Yes, in part. To find the moments, only the sound of your video (a light MP3) is sent to TonySkill's server, transcribed, then deleted right after processing. The video itself is only sent if you ask for the MP4 clips (or if your browser can't read the file), or downloaded by the server if you paste a YouTube link; it is deleted as soon as you are done and in any case within 45 minutes. Nothing is shared with anyone or kept.

How does the AI pick the moments?

An open language model (Qwen3, Apache-2.0) reads the whole transcript and looks for complete moments — a strong question, a surprising idea, an anecdote — that fit the length you asked for. It also suggests a title and a caption. It's a suggestion: watch each moment before posting.

Is it as good as OpusClip?

For picking moments and preparing the vertical format, the result is close. Paid tools add a virality score, animated emojis or multi-person reframing; here, subject tracking works well on one person talking, less on a moving group.

Which videos work?

Any video with speech: podcast, interview, talk, vlog, lesson, up to 45 minutes and 2 GB (MP4, MOV, WebM, MKV), or a YouTube link. A video without speech (music, scenery) gives nothing: the AI works on what is said.

Related tools

  • Resize video (9:16, 1:1…)Switch a video to 9:16, 1:1 or 16:9 for TikTok, Reels or YouTube: blurred background, bars or crop.FreeNo sign-up
  • Automatic subtitlesCreate SRT or VTT subtitles for a video or audio file, fix them line by line, with nothing uploaded.FreeNo sign-up
  • Add subtitles to videoTikTok-style word-by-word animated captions, free text or an .srt file burned into the video.FreeNo sign-up
  • Trim videoShorten a video to the part you want, to a tenth of a second, or instantly without re-encoding. MP4, MOV, WebM.FreeNo sign-up
  • Video editorEdit videos in your browser: trim, join, speed up, 9:16 TikTok format, text, subtitles, music.FreeNo sign-up