Import one talking-head video. Studio Editor transcribes what you said, writes word-timed captions in that language or translates them into another, adds matching stock footage, emoji and stickers, and opens the result in an editor where you can change anything before exporting a finished 9:16 video to your phone.
Free to start 69 spoken and caption languages Video file never uploaded
Everything below is in the app today, on iPhone and Android. Each automatic step can be switched off or edited afterwards.
Studio transcribes your speech with word-level timestamps and turns it into short, readable captions. The text is kept verbatim in the language you spoke; nothing is paraphrased or summarised. Only words the transcriber flagged as uncertain are offered for repair.
Captions are burned into the exported video, so they show on every platform without relying on the platform's own subtitle feature.
Tell Studio which of 69 languages is spoken in the video so transcription is accurate. English sits in the same list as Nepali, Hindi, Spanish, Arabic and the rest; there is no "supported languages" hierarchy.
Turn on Translate captions and choose a caption language that differs from the spoken one. Speak Nepali and caption in English, speak Hindi and caption in Arabic, speak English and caption in Japanese. Any of the 69 languages can be the source or the target.
For languages not written in Latin letters, choose whether captions appear in the native script or a romanised version, for example कहाँ जाने or kaha jane. Romanising is not translating: the words stay the same and only the letters change, which suits creators whose audience reads the language but types it in Latin letters.
With Smart B-Roll on, an AI model reads the transcript, works out what each moment is about, and searches licensed stock footage for a matching clip. Cutaways are placed where the words carry the meaning, kept short, and capped so the video never turns into a stock montage. The voice continues underneath every cutaway.
Auto Emojis adds contextual emoji to captions. Animated icon stickers and a light transition sound are placed at key moments. Emoji can be toggled off in setup or in the editor; stickers are part of Smart B-Roll and switch off with it.
Type a one-line summary such as "travel vlog in Pokhara" or "gym workout tips". It helps the AI spell names and places correctly and pick relevant footage. It never changes words that were transcribed confidently: the transcript is what you said, and the transcript wins.
Studio Editor shows a live preview above a three-track timeline: B-roll overlays, your footage, and captions. Tap a caption to rewrite it. Drag its handles to retime it. Move, trim or delete a B-roll clip. Switch caption template, colour, position and size. Every change is undoable and nothing is re-encoded until you export.
Export renders once, on your phone, using the hardware encoder. The output is a 9:16 vertical video at the resolution of your source (1080p, 1440p or 4K where your device supports it), saved to your gallery with share shortcuts for TikTok, CapCut and Instagram.
Pick one for the whole video in the editor's Templates tab; text colour and font are adjustable too.
Five steps. The AI does the transcription, writing and footage search; you make the choices and the final call.
On the home screen, tap Auto Captions & B-Roll and choose one video from your gallery. Studio is built for a single talking-head clip: you speaking to the camera.
Under Language & captions, set Spoken in to the language of the video. To caption in a different language, switch on Translate captions and choose Translate to. For non-Latin languages, pick native or Roman script.
Studio extracts the audio on your phone, transcribes it with word-level timing, writes the captions (translated if you asked) and searches for matching footage. A stage-by-stage progress bar shows what is happening. If the AI planning step fails for any reason, you still get a captioned video from the transcript.
The finished short plays first. Tap Edit to open Studio Editor if you want changes.
Tap Export. The video renders once on your phone and is saved to your gallery, with shortcuts to open it in TikTok, CapCut or Instagram. Free exports carry a small watermark; Pro exports do not. Your project is kept as a draft so you can come back and re-export.
Every language below can be the spoken language, the caption language, or both. Languages marked with their native name also offer a Roman-script option.
If your video is you speaking, Studio Editor removes most of the editing that stands between recording and posting.
Post captioned TikToks, Reels and Shorts without opening a timeline editor. Captions, B-roll and emoji in one pass, then fix the odd word.
Turn a piece-to-camera into a proper edit: cutaways where you describe something, captions that keep viewers watching on mute.
Describe a place and let Smart B-Roll find footage of it. Caption in your language, or translate for an international audience.
Accurate, verbatim captions for lessons, tutorials and tips, with illustrative footage where it helps and none where it would distract.
Speak Nepali, Hindi, Bengali, Urdu, Tamil or any of 69 languages and reach viewers who read another. Native or Roman script, your choice.
You do not need editing experience. Import, pick a language, generate, export. The editor is there if you want it, not required.
Studio Editor is designed so the automatic result is usually right, and every part of it is yours to override.
| Element | Done automatically | Editable in Studio Editor |
|---|---|---|
| Transcript | Word-level transcription in the selected spoken language | Rewrite any caption's text |
| Caption timing | Captions cut at natural phrase boundaries, aligned to the words | Drag handles to retime; move or delete |
| Caption language | Verbatim by default; translated when you choose a different caption language | Set in setup; edit individual translated lines afterwards |
| Script | Native script by default for non-Latin languages | Choose native or Roman in setup |
| B-roll | Stock footage matched to meaning, placed and trimmed to fit | Move, trim the start or end, or remove any clip; turn the feature off |
| Emoji and stickers | Contextual emoji in captions, icon stickers at key moments | Toggle emoji on or off for the whole video; stickers are governed by the Smart B-Roll switch in setup |
| Caption style | Default Bold style | Choose one of ten templates; set colour, font, position and size |
| Export | 9:16 at source resolution, hardware-encoded on your phone | Re-export any time from the saved draft |
What Studio does not do: it does not add music or a voiceover, and it does not combine multiple clips into one video. Add music in the app you post from, and use Multi-Clip Reel when you have many clips rather than one.
Real screens from the app. Swipe to scroll.
Swipe or scroll →
Studio Editor is the AI video editing tool inside the KatKut AI app for iPhone and Android. You import one talking-head video, choose the spoken language and the caption language, and it generates word-timed captions, optional translation, matching stock B-roll, emoji and stickers. You can edit every caption and clip in a timeline editor, then export a 9:16 vertical video to your gallery.
Studio extracts the audio from your video on your phone and sends only that audio to a speech-to-text service, which returns a transcript with a timestamp for every word. An AI model then splits the transcript into captions, adds emoji where they fit, and, if Smart B-Roll is on, chooses short English search terms that describe what each moment is about. Those terms are used to find licensed stock footage, which is downloaded to your phone. Your own code on the device checks every caption and cutaway for timing and overlap, then shows you the result before anything is rendered.
Open KatKut AI, tap Auto Captions & B-Roll, pick a video, set the spoken language, and tap Generate. Captions are created automatically and burned into the video when you export. You can change the caption style, colour, size and position in the editor before exporting.
On the Studio setup screen, tap Spoken in and choose from the list of 69 languages. You can search the list by name. The language you pick is passed to the transcriber, so choosing the right one is what makes the captions accurate.
On the setup screen, switch on Translate captions and choose a Translate to language. The spoken language and the caption language are set independently, so any of the 69 languages can be the source or the target. Translated captions can still be edited line by line in the editor.
Yes. Nepali is one of the 69 spoken and caption languages. You can caption Nepali speech in Devanagari or in romanised Nepali, translate Nepali speech into English or any other supported language, or translate speech in another language into Nepali captions.
With Smart B-Roll turned on, the AI reads the meaning of what you said and searches for licensed stock footage that illustrates it, for example footage of a street market when you talk about one. Clips are kept short, placed only where the words carry the meaning, and limited so they never cover more than half the video. Your voice continues under every cutaway. You can move, trim or remove any clip in the editor, or turn the feature off entirely.
Yes. Studio Editor has a live preview and a timeline with three tracks: B-roll, your footage and captions. Tap a caption to rewrite its text, drag its handles to change when it appears, or delete it. Drag a B-roll clip to move it, trim its start or end, or remove it. Choose from ten caption templates and set the text colour, position and size. Undo and redo cover every change. Nothing is re-encoded while you edit; the video is rendered once, when you tap Export.
Tap Export in the editor or on the result screen. The video is rendered on your phone using the hardware encoder and saved to your device gallery as a 9:16 vertical video at your source resolution. Share shortcuts for TikTok, CapCut and Instagram appear when it finishes. Free exports include a watermark; Pro exports do not.
Yes. KatKut AI is on Google Play for Android and on the App Store for iPhone (iOS 16.4 or later). Studio Editor is included on both platforms.
On the free plan, Studio captions up to 4 minutes of video per month, each video up to 1 minute long, with a watermark on exports and ads before and during export. All languages, translation, B-roll and caption styles are included. KatKut AI Pro costs $3.99 per month and raises the allowance to 120 minutes per month and 5 minutes per video, with no watermark and no ads. The local price may vary on the App Store. Pro is purchased from Settings in the app: through Apple's in-app purchase on iPhone, and through a secure web checkout on Android.
Your video file is never uploaded. Studio extracts the audio track on your phone and sends only that audio to a transcription provider. The text transcript, plus the optional topic you typed, is sent to an AI model to write captions and choose B-roll search terms. Search terms alone are sent to the stock footage providers. Neither the audio nor the transcript is stored on our servers. Rendering and export happen entirely on your phone. The full list of providers is in the privacy policy.
One clip of a person speaking to the camera, recorded in a reasonably quiet place, in a single language. Studio takes exactly one video per project. If you have many short clips rather than one, use the Multi-Clip Reel tool instead.
Not inside Studio. It keeps your original voice and adds captions, B-roll, emoji and light sound effects. Add music or a voiceover afterwards in the app you post from, such as CapCut, TikTok or Instagram, using the share shortcuts on the export screen.
More general questions about the app, accounts and billing are answered on the About page.
Download KatKut AI, open Studio Editor, and caption your first video in the language you speak. Free to start.
Requires iOS 16.4 or later on iPhone. Studio needs a free account so your monthly minutes can be tracked.