How to Add Subtitles to a Video

To add subtitles to a video, upload it, ask, and download two files a few minutes later: the video with subtitles burned into the picture, so they show wherever it plays,, and a matching SRT for platforms that take a caption file. Styling is one more sentence.

Attach the video and say what you want

Add your video to a new chat, any common format, and say what you want: "Add subtitles to this video" is enough. Chat Octopus transcribes the speech and times each word to the moment it's spoken.

Two files come back:

• The subtitled video: an MP4 that comes back in the same shape as the video you sent, with the subtitles burned into the picture so they show wherever it plays.
• A matching SRT: the same words, timed the same way, saved as a separate caption file for platforms that take one instead.

On screen, lines run about four words at a time. The SRT groups the same speech into lines of up to eight words.

Position only applies to the burned-in video: an SRT carries no coordinates of its own, so wherever it lands on screen is up to the app or player showing it. On the video, the default line sits a short distance above the very bottom of the frame, clear of the buttons and icons every app plants along the edge, on any shape of video. That spacing is built for vertical, so the gap is a fixed size, so the shorter the frame, the further up the line sits: about a third of the way up on widescreen. Ask for a different band and the next version moves there.

Starting from raw footage rather than something already cut? Ask the footage which moments are worth keeping first, then caption the clip once it's cut.

Describe the style you want

There's no style menu to click through. Say what you want changed and the next version has it. Leave it alone and you get the built-in look, white text on a soft dark box behind the words, with the word being spoken picked out in yellow as it's said, then back to white once the next one starts. That's karaoke-style timing: one word lit at a time.

Font, color, size, the box behind the text, and the animation are all things you describe. Position is the one limit worth knowing. Three bands only, bottom, center, or top, with the text centered inside whichever one you pick, never dropped into a corner or nudged to an exact pixel. A tutorial captioned over a busy screen recording often wants a black outline instead of the box, so the text still reads against moving detail behind it.

One habit worth keeping: stick to a single accent color. One highlighted word per line stands out, three competing colors just fight the footage for attention.

Fix mistakes by talking

Captioning only has the sound to go by, so it stumbles on the same things a person would with headphones on and no video: two people talking over each other, a strong accent, background noise sitting under the words, a brand or product name it has never heard before. Watch the result once, and when a line is off, send the fix as its own sentence. Ordinary words come back clean; it's your own vocabulary, the founder, the product, the company, that's worth one deliberate check before you sign off.

• "The company name in the second line should read Nordic Sound, not 'nordic sound.'"
• "The last subtitle cuts off early, stretch it to the end of the sentence."

Only that line changes. The transcript and the styling both stay attached to the thread, so the fix lands on the line you flagged. The video does get built again around the fix, but that rebuild is machine time, not yours.

Add another language

Once the first set of captions reads the way you want, ask for a second language in the same thread: "Translate these into French, same styling." Both versions sit in the conversation afterward, so a fix to either one doesn't mean starting over. For a fix that lands later, just note which language it's in.

Get them onto YouTube, TikTok, or Reels

There is no direct upload to any platform: you download the finished files and post them yourself. Each platform gets the file it handles best.

• TikTok, Reels, Shorts: post the burned-in MP4. These feeds autoplay on mute, and burned-in subtitles always show, styled exactly the way you asked.
• YouTube: upload your video as usual, then add the SRT to it in YouTube Studio's subtitle settings. Viewers can turn the captions on and off and read them at their own size.
• A platform that wants a different caption format: ask for VTT, built for web-based video players, or ASS, which bakes the styling into the file itself and always comes back bottom-centred, and export that instead.

As a rule, sound-off feeds want the burned-in file; anything with its own caption toggle wants the SRT instead. Captions and subtitles are different jobs if you want the fuller reasoning behind that split.