Driven by the assistant
It cuts, titles, captions and changes speed, then reports the result.
AI video editorIn closed testing
An AI video editor should make the edit, not only suggest one. The KolHive video editor is one of the apps on the desktop, and the assistant can work its timeline: cut, add titles and captions, change speed and set up Face & body retouch, then read back what changed. Underneath is a full editor, with a multi track timeline, caption tracks in the formats broadcasters ask for, motion titles, chapters and sound tools.
It cuts, titles, captions and changes speed, then reports the result.
Several tracks per project, in SRT, WebVTT, SCC, TTML, DFXP, EBU-TT-D and EBU STL.
Thirty built-in presets for titles, lower thirds, call-outs and end cards.
Face & body retouch runs locally. Its models download the first time you use it.
A subtitle file is only useful if the place that receives it can read it. Each caption track keeps its own language and style and exports on its own.
Keep a transcript in one language and translations in others on separate tracks of the same project.
Import and export SRT, WebVTT, SCC, TTML, DFXP, EBU-TT-D and EBU STL, for web players, social platforms and broadcast hand-offs.
EBU STL files can be written for Latin, Cyrillic, Arabic or Hebrew text, and the editor itself mirrors for Hebrew and Arabic.
Burn one track into the picture, or export the caption files next to the video.
The finishing work that usually sends a project to three other tools stays on one timeline.
Titles, lower thirds, call-outs, social buttons, end cards and counters from thirty presets, in the colours you choose.
Ramp the speed of a clip, play it backwards, hold a frame, and choose frame blending.
Mark moments with notes and ranges, then write them as chapters in MP4, MOV and M4A files or as a YouTube chapter list.
Fit a piece of music to the length of the edit, and level clips to a loudness target.
Switch between Edit, Color, Audio, Captions and Effects layouts. Each one remembers its own panels on that device.
Smooth skin and adjust the face and body of the people in a clip, with the strength animated over time. It runs on the device.
Add the video and its audio to a project, or open a recording from KolWrite.
"Remove the long pauses", "speed up the intro": the assistant edits the timeline and tells you what it did.
Place motion titles, then add a caption track for each language you need.
Export the video, with captions burned in or as separate files in the formats the destination accepts.
Yes. The assistant can cut, add titles and captions, change speed and set up Face & body, and it reads back the result so you can see what changed.
SRT, WebVTT, SCC, TTML, DFXP, EBU-TT-D and EBU STL, each as its own caption track.
No. It analyses and renders on your device. Its models download the first time you use it and can be removed in Settings under Storage.
Yes. The KolWrite API returns transcripts with word timings and SRT files that an agent can use to cut video. The documentation is at console.kolwrite.com.
KolHive is in closed testing. For subtitles today, KolWrite makes SRT and VTT files from your videos.
KolHive is in closed testing. Join the launch list and we will tell you when you can open it on your own machine.