What's New in Skapo.
Follow along as we ship. We post updates every week, making the clips faster, sharper, and easier to create.
Captions That Stay Off the Face, and a Fix Button
Two things that were costing people a re-render. Captions now measure where the speaker's face actually sits in each clip and place themselves under it instead of on the chin. And when a word comes back wrong, you fix it in the browser and re-burn that one clip, rather than spending credits on the whole video again.
- •Smart caption placement measures the speaker's face band per clip and drops the captions below it. On by default for Freelancer Pro and White Label.
- •Your position always wins. Drag the caption slider, or pin a position in a client's brand kit, and Skapo leaves it exactly where you put it.
- •It is skipped where it cannot help: split-screen layouts, and most 16:9 output, where the fixed chest-zone height is already right.
- •The clip editor is text-first. Cut a sentence, trim the ends, reorder, split at the playhead, or fix a mistyped word, then re-burn that one clip. Every save is a version, and you can restore an older one.
- •Caption files come out on their own too: SRT, VTT or TXT per clip, plus the full episode transcript.
- •Clips also come out as timeline files. Each one downloads as an FCP7 XML, which Premiere Pro imports as Final Cut Pro XML and DaVinci Resolve reads too, or as a CMX3600 EDL, and it rebuilds the clip as cuts against your original video rather than handing you our finished MP4.
- •One XML per video holds every clip as its own sequence against a single media reference, so you relink the episode once and the whole bin comes online. Captions are not inside the file: import the SRT next to it, the way editors already do.
Captions That Render in Asian Scripts
Skapo could already transcribe Japanese, Chinese, Korean, Thai and Hindi. It just could not draw them: every caption font we shipped was Latin-only, so those languages burned in as rows of empty boxes with perfectly correct timings and no error anywhere. That is fixed. Seven languages moved from transcribed-but-unusable to fully supported.
- •Japanese, Chinese, Korean, Hindi, Thai, Indonesian and Malay are now tuned end to end, taking the curated list from 15 languages to 22.
- •Captions pick their typeface from the script of the text itself, so Japanese gets a Japanese face, Korean gets the only face that draws Hangul, and Cyrillic and Greek stop falling back to a font that never had those letters either.
- •Japanese, Chinese and Thai lines are now sized to fit the frame. These languages are written without spaces, and the subtitle renderer only ever breaks a line at a space, so a long line used to run straight off both edges of the video.
- •Filler-word trimming and clip-opening rules are tuned per language rather than borrowed from English.
- •Language detection no longer suggests only European answers, which had been quietly biasing it toward the wrong code on Asian audio.
The First 3 Seconds: Viral Hook Engine
The opening of a short decides whether anyone watches the rest, so we rebuilt the first three seconds of every clip. Hook title cards are now on by default for everyone, the opening is never a static talking head, and your strongest hooks get made first. Every short also gets its own cover image for a scannable library.
- •Bold scroll-stopping hook titles are now baked onto the opening of every clip, on every plan, not just Pro.
- •New Viral Motion: a slow push-in over the hook, alternating punch-ins between cuts, and a tightened loop, so the first seconds always feel alive and the clip loops cleanly back to the hook.
- •Clips are now ranked by hook strength before rendering, so your most scroll-stopping moments are produced first.
- •Every short gets an auto-generated 9:16 cover thumbnail (the best hook frame with your title on it), turning your library into a scannable wall of clips, built for agencies reviewing dozens at a glance.
- •Render timeouts now scale with video length, so long podcasts and interviews get the processing time they need.
Hormozi-Style Animated Captions
Skapo now has a built-in Hormozi caption style, the bold, word-by-word, high-contrast look that's everywhere in short-form for educational and business content. No plugin, no fiddling with templates. Pick it once and it's on every clip.
- •Words pop on one at a time, with your own highlight color, shadow, and all-caps option, set to the classic Hormozi look out of the box.
- •The caption outline adjusts itself so text stays readable on both dark and bright footage.
- •Each client can have their own caption style, one gets Hormozi, another gets minimal, all in the same batch with no manual switching.
- •Captions always sit where the TikTok, Reels, or Shorts buttons won't cover them.
Everyone Stays in Frame
Panels and interviews used to crop down to one person and lose the rest of the room. Skapo now finds the faces in the wide shot and builds the vertical frame around them, so a two, three or four person conversation still reads as a conversation.
- •FaceTracker spots the people on screen and stacks up to 4 of them into their own panel of the 9:16 frame, picking the layout that fits: a dual stack, a top pair over a full-width row, or a 2x2 grid.
- •On a single-speaker crop the frame follows whoever is talking, so the camera is never parked on an empty chair while someone off-center answers.
- •To be clear about what this is not: Skapo does not label who said what. There are no speaker names on the transcript and no searching by speaker.
- •Framing reads the picture, not the words, so it behaves the same across all 22 tuned languages and the 40+ that get word-level captions.
Silence Gap Detection & Automated Dead Air Removal
Clips produced by Skapo now automatically strip silence gaps, long pauses, and dead air from the final output. The result is tighter pacing without manual timeline work, a major time save for agencies processing high-volume podcast and interview content.
- •Set how long a pause has to be before it's trimmed (default: 0.4 seconds), with a different setting per client, some content needs room to breathe, some doesn't.
- •Tells the difference between a meaningful pause (emphasis, a dramatic beat) and plain dead air, and only trims the dead air.
- •Your captions stay perfectly in sync after the cuts, nothing drifts out of place.
- •Average clip runtime reduction of 8-14% on unedited podcast source material, directly improving completion rates.
Support for 15+ European Languages & Automatic Language Detection
Skapo now understands fifteen European languages and detects which one you're speaking automatically, keeping each speaker's flow intact and trimming filler words cleanly.
- •Auto-detects Danish, Dutch, English, Estonian, Finnish, French, German, Italian, Latvian, Lithuanian, Norwegian, Polish, Portuguese, Spanish, and Swedish.
- •Trims filler words ('uhm', 'eh', 'ehm') cleanly, tuned for each language, not just English.
- •Captions now get punctuation right across all the supported languages.
Secure Direct Video Upload
You can now upload big video files straight to our own private, encrypted storage, they never pass through anyone else's servers.
- •Checks your video right in your browser before it uploads, so bad files get caught early.
- •Shows you the credit cost and any file-size fee up front, before you commit.
- •GDPR-friendly: everything is stored encrypted and automatically deleted after 7 days.
Editor Twin & Smarter Hook Detection
Big updates to Editor Twin, the AI that finds your best moments, with a cleaner look and sharper picks for which openings will grab attention.
- •Rates how strong each opening hook is and surfaces your video's highlights automatically.
- •Cuts on the natural pauses in speech, so clips never get chopped off mid-word.
- •Places captions where the YouTube Shorts and TikTok buttons won't cover them.
Want to stay updated automatically?
Get early access to new features before anyone else by joining our waitlist today.