Two kinds of captions, and why you might want both
There are two ways to put captions on a YouTube video:
- Burned-in (open) captions are part of the picture. Everyone sees them, on every platform, with the sound off. This is the animated, word-by-word style you see on Shorts, TikTok and a lot of talking-head videos.
- Closed captions are a separate file, usually SRT, that you upload in YouTube Studio. Viewers can turn them on or off, YouTube can translate them, and the text is attached to the video as a caption track.
Many creators use both: burned-in captions to keep people watching with the sound off, and an SRT so the caption track is accurate instead of YouTube's automatic one. Studio gives you both from one upload.
How YouTubePilot adds captions
Studio transcribes your video word by word, with a start and end time for every word. It detects the language on its own, or you can set it. Then it groups the words into short lines of up to three words, breaking at the ends of sentences, and renders them into the video.
Because the same pass also cuts pauses and filler words, the captions are timed to the edited video, not the raw take. You don't have to re-sync anything after cutting.
There are three caption styles:
| Style | What it looks like | Good for |
|---|---|---|
| Pop | Large bold words with an outline. The word being spoken lights up in a highlight colour. | Shorts, fast talking-head videos |
| Clean | Plain white lines with an outline, smaller and calmer. | Tutorials, longer videos |
| Boxed | White text on a black box. | Busy backgrounds, screen recordings |
The highlight colour can be yellow, green, red, cyan or white. Captions are in capitals by default; you can switch to normal case. You can also turn burned-in captions off and only take the SRT.
Vertical video gets captions sized for a phone screen, and the Shorts Studio clips from your video are captioned the same way.
What you get back
- The edited MP4 with captions burned in, up to 1080p.
- An SRT file timed to the edited video, ready to upload in YouTube Studio under Subtitles.
- A plain-text transcript.
- Five title options, a description with chapters, tags and thumbnail text.
- Up to five vertical Shorts with captions.
How to add captions to a video in 3 steps
- Upload the video. Open Studio and drop in your file.
- Pick a caption style. Pop, clean or boxed, and a highlight colour. Choose how tight the pause cuts should be, or turn them off.
- Download. Take the captioned MP4 for upload, and the SRT if you want a closed-caption track too.
Don't like the style? Render again with a different one. It reuses the transcript, so it doesn't use more of your minutes.
Accuracy, honestly
Automatic transcription is good, not perfect. Names, brand names and niche jargon are where it slips most. Read through the transcript before you publish, especially on a video where one wrong word would be embarrassing. The SRT is plain text, so it is easy to fix a word before uploading it to YouTube.
From the terminal, Claude Code or Cursor
Studio is also a command-line tool and an MCP server, so you can run it from a script or hand it to an AI agent. Install the CLI, save your API key, and edit a file:
curl -fsSL https://youtubepilot.com/install.sh | shytpilot login ytp_...ytpilot edit video.mp4 --captions boxed --highlight green
The CLI is a single Python file with no dependencies. Add --json and it prints a
machine-readable result, which is what agents need.
If you work in Claude Code, Cursor or another MCP client, add the server instead and ask in plain words, for example "edit ~/Desktop/take3.mp4 with razor pauses and boxed captions":
claude mcp add --transport http youtubepilot https://youtubepilot.com/mcp --header "Authorization: Bearer ytp_..."
The MCP tools can upload a local file or fetch one from a direct link (Dropbox, Google Drive, S3), check progress, return the titles, description and download links, and render an edit again with different settings. API and MCP access come with every plan, Free included.
Free gives you 15 minutes of editing (with a small watermark and one Short) plus one niche analysis, no card. Paid plans remove the watermark and add Shorts, the API and MCP: Starter is $19 a month for 2 hours of editing, Creator $39 for 8 hours, Studio $79 for 25 hours. See pricing.