Create a caption file for your video, watch it play over the clip, then attach it on the platform where the video lives. Nothing is uploaded, and there’s no account or watermark.
Choose the video you want to caption (MP4, MOV, WebM or MKV). Set “Spoken language” if you know it, which is quicker and more reliable than detection.
Press “Generate subtitles” and wait for “Transcribing…” to reach 100%. Captions fill the list while the model works.
Play the clip in the “Preview” panel. Click any subtitle to jump to that moment, fix the wording, and adjust “Start” or “End” if a line shows too early or too late.
Click “Download SRT” (or “Download VTT” for web players) and upload that file next to your video on YouTube, in VLC or in any app that accepts subtitle files.
There are two ways to put captions on a video. Soft subtitles, also called closed captions, live in a separate file (usually SRT or VTT) that the player draws on top of the picture. Viewers can turn them on or off, choose a language, and the platform can index the words. Hardcoded subtitles, or open captions, are painted into the frames themselves and can never be switched off.
filetidy makes soft subtitles. You get an SRT or VTT file that you add to the video where it is published: in YouTube Studio, open Subtitles, pick the video, choose Add language, then upload the file with timing. In VLC, use the subtitle menu to add the file. Many editing and social apps also have a caption import that accepts SRT.
Burning captions permanently into the picture is not available in filetidy yet; it is planned. Until then, a soft subtitle file is the more flexible choice anyway: you can fix a typo later without re-exporting the whole video.
Recognition runs locally with the Whisper model, so the clip you are captioning never leaves your computer or phone, even if it is unpublished.
Mac
The subtitles are shown while playing in VLC; the video file itself is not changed.
Windows
Clipchamp requires a Microsoft account and processes the video in its own service. Caption languages and export options can change between versions.
SRT is the safe default. YouTube, Facebook, LinkedIn, Vimeo, most video editors and desktop players such as VLC import it. VTT is what browsers use for videos embedded on a website, so pick it if you are adding captions to your own page or a web course platform that asks for WebVTT.
Uploading captions as a file usually beats burning them in. Search engines and platforms can read the text, viewers who don't need captions can hide them, and you can upload several languages for the same video. Burned-in captions make sense mainly for short social clips that autoplay without sound, on apps that don't accept a subtitle file.
Before uploading, watch the whole clip once in the preview. Timing problems are easy to spot there and much harder to fix after the video is live.
Not yet. Today filetidy produces a subtitle file and shows it over your video in the preview, but it does not render the text into the frames. Hardcoded captions are planned for a later version.
In YouTube Studio, open Subtitles in the left menu and select your video. Add the language if it is not listed, choose to add subtitles, pick Upload file, select With timing, and choose the SRT you downloaded. YouTube shows a preview before you publish.
Speech recognition places each line from the timestamps it predicts, which can be off by a fraction of a second, especially after music or long silences. Click the line in the list, then edit its “Start” and “End” times while watching the preview.
Open “Line layout”, lower “Max characters per line” or switch “Lines per subtitle” to 1, then press “Re-split subtitles”. The captions are rebuilt with the new limits and your text corrections are kept.
No. A soft subtitle file sits beside the video; the picture and sound are untouched. Players and platforms read the two together, and you can replace the subtitle file at any time.
The free on-device version captions the first 10 minutes of each file. For longer videos, a Pro version that handles files of several hours is coming; its waitlist will open on the pricing page.
Last updated October 28, 2026
Choosing the language is faster and more accurate than detection, especially for short clips.
Speech recognition makes mistakes, especially with names, numbers, music and several people talking at once. Dialects such as Moroccan Darija are recognised less reliably than Modern Standard Arabic. Read the subtitles before you publish them.
Need long videos or the best accuracy?
AI subtitles Pro will use Whisper large-v3 on our servers: about an hour of video in a few minutes, better accuracy on accents and dialects, files up to 3 hours. Only compressed audio is sent, never the video.
See Pro and join the waitlist