How it works
From a folder of raw takes to a cut you can review, in four steps.
Step 1
What you upload
A project is one video. Put everything for it in one place — Takefold sorts it out.
- Every take, as many as you shot — several takes of the same line are expected.
- The client's script. Each line becomes a segment of the cut.
- Your notes or the client's brief, as text or a PDF: the tone, what to avoid, where to use cutaways.
- A music track, if you have one. If not, Takefold can suggest a Creative Commons track that fits.
- Upload from your computer or import from Google Drive.
Step 2
What the AI looks at
Takefold watches and listens to every take before it chooses anything.
- Every word, with its timing, in English or Brazilian Portuguese.
- Pauses, breaths, filler words and restarts.
- Framing, focus and exposure, and whether the speaker is looking at the lens.
- Clips with no speech are kept as b-roll, used only where your notes ask for cutaways.
- For each line it picks the take that says it whole, without stumbles, and trims it with room to breathe — never in the middle of a word.
Step 3
What you can change
Nothing is rendered until you say so. The proposal is a draft you can take apart.
- Swap a take for one of the alternates, or leave a line out.
- Move a cut earlier or later.
- Edit the captions' text and timing over the player, and pick a caption look.
- Keep the suggested track or choose one of the runners-up.
- Ask the editor in the chat: “why this take?”, “use the second take for the opening”. A bigger change becomes a new version.
Step 4
What you get
When it looks right, build the video.
- A video in the format you shot — vertical takes give a vertical video.
- Captions burned in, and the same captions as an .srt file.
- Music under the voice, faded in and out, and loudness evened out.
- Every render is kept as a version, so you can go back.
Try it with your own takes
Takefold is free during early access. Create an account and start your first project.