Content Studio
Nothing posts without your approval.
What do you want to make?
Mikey chat-to-cam type a script and Mikey says it
to camera. The voice is made first, then the video. Watch it, then save it to the
library with its script.
0 characters
Results newest batch first · cost, wall-clock and
model on every tile
Describe what arrived the prompt is what we asked
for. This is what the clip actually shows, and it is the only thing search looks
at.
Clip library everything we have made. Search looks at
what the clip ACTUALLY SHOWS, never at the prompt.
Music library tracks you own, stored here for
reference and playback. No model takes audio, so nothing here is attached to a
generation.
Drop tracks here
MP3, WAV or M4A. Up to 40 MB a track.
Every file is checked by its contents, not by its name.
Mikey voice type a script, get Mikey reading it.
Every take is saved with its script and settings, so any of them can be made
again exactly.
0 characters
Drop a video here
Or click to choose one. Full length is fine — up to 10 GB per video.
Pick the brand for this video before uploading. Default is no logo — better none than the wrong one.
On by default. This is the run's default — you can still turn captions off on
an individual clip later. The first-8s hook stays either way; clear the hook text itself to drop it.
Processing
starting
Every video processed
Loading...
Clips
Upload a video, or pick a run under Runs, to see its clips.
Content type:
Proposed 9:16 crop on every clip. Spot a badly framed one, nudge it, then re-render just that clip.
Who sits where — confirm before rendering speaker-cut clips
This map applies to every clip in this episode. Check each seat's face matches its speaker; low-confidence seats are flagged. Fix any wrong one, Save, then render.
Posts
Pick a run under Runs to see and edit its copy.
Mids — topic segments
The episode cut into 3–10 minute topic segments covering the whole
thing, in its natural form: original landscape, no crop, no burned captions, no hook.
Mids do not go through the Queue. Buffer treats every YouTube video post as a Short
(max 3 minutes, portrait) and rejects long-form, so a mid is downloaded here and uploaded
in YouTube Studio — where the thumbnail has to be attached by hand anyway. Its title
and description are written and approved under Posts, ready to paste.
Photo posts
Type the hook that will be burned onto the image. The library is matched
against it and the five best photos are rendered as 1080×1350 posts.
Options
One photo, chosen by you
Pick the photo yourself and the matcher is skipped entirely. Everything
after that is identical — same pill, same character overlay, same nudge, same
approve and queue. One option instead of five.
Library
Loading…
Buffer queue
Pick a run under Runs to see what's ready to send and what's already been queued.
Approval gaps — across every run
Preview