How does the caption generator work?

Upload your video or audio and paste the exact words -- a script or song lyrics. We transcribe the audio only to get the timing (when each word is spoken or sung), then snap your text onto that timing and hand you a downloadable SRT subtitle file. Your words stay exactly as you wrote them; we just line them up to the audio.

Does it work for speech, or only songs?

Both. It works for any spoken audio -- talking-head videos, UGC ads, podcasts, interviews, voiceovers, faceless videos -- and for sung songs and lyric videos. If you have the words and the audio, it times them.

Does it work in languages other than English?

Yes -- any language. This is the whole point. CapCut's auto-captions and most auto-subtitle tools only handle a handful of languages well and mis-hear the rest. Because you supply the real words, the captions are always correct in Spanish, French, German, Portuguese, Polish, Arabic, Hindi, Japanese, Korean, or any other language. We only figure out the timing.

Will the words be spelled correctly?

Always. Auto-caption tools guess the words from the audio and get brand names, proper nouns, slang, and sung word-endings wrong. Here your pasted text is the source of truth, so every word -- and every accent and diacritic -- is exactly what you typed.

Can I use the SRT in CapCut, Premiere Pro, or DaVinci Resolve?

Yes. SRT is the universal subtitle format. Import it straight into CapCut, Adobe Premiere Pro, DaVinci Resolve, Final Cut Pro, VN, InShot, or any editor and every line lands on the right frame. You can then style it however you like.

How is this different from CapCut auto-captions?

Auto-captions transcribe the audio and guess the words -- which breaks on non-English audio, brand names, and singing. This tool flips it: you provide the exact words and we only solve the timing. So it works in every language and never misspells, which auto-captions can't guarantee.

Do I have to type any timestamps?

No. You never touch a timestamp. Paste your words one line per caption, and every start and end time is calculated for you from the audio.

What if my script or lyrics don't match the audio word-for-word?

That's fine. The alignment tolerates small differences -- an extra ad-lib, a repeated word, a dropped filler. It anchors the words it recognizes and evenly spaces the rest, so the timing stays smooth.

Vertical or horizontal -- what's the difference?

It controls the line length so captions fit the frame. Vertical (9:16) for TikTok, Reels, and Shorts caps each line at 25 characters; horizontal (16:9) for YouTube and TV caps at 42. You can switch between them after generating with no re-processing.

What files can I upload?

MP3, WAV, or M4A audio, or an MP4 video. Audio files up to 50 MB and videos up to 200 MB.

Is it free, and do I need an account?

It's completely free and needs no signup. Upload, paste your words, and download the SRT.

Free Speech & Song Caption Generator

Turn any video or song into perfectly timed SRT captions from your own script or lyrics. Works in any language, keeps your exact spelling, and imports straight into CapCut, Premiere Pro, or any editor.

Make the whole ad, not just the captions

Starpop turns your product into scroll-stopping video ads with AI -- in minutes.

Create with Starpop

1. Your video or song

Drop your audio or video here

or click to browse — MP3, WAV, M4A, or MP4

2. Your exact speech or lyrics (optional)

Leave this blank and we'll transcribe the audio for you. Paste your script or lyrics — one line per caption — to keep every spelling exactly, in any language. Either way we handle the timing.

3. Video format

How it works

Upload your audio or video

MP3, WAV, M4A, or MP4 — a song or any spoken clip.

Paste your exact words

Your script or lyrics, one line per caption, in any language.

Download the timed SRT

Import into CapCut, Premiere Pro, or any editor. Timestamps done for you.

Captions in any language

Auto-caption tools like CapCut's break on most languages and mishear brand names and sung words. Because you paste the real text, this generator produces correct subtitles in every language — accents and diacritics intact — and only solves the timing.

EnglishSpanishFrenchGermanPortugueseItalianPolishDutchTurkishArabicHindiJapaneseKoreanMandarinVietnameseIndonesian+ any other