Caption generator – automatic subtitles as SRT, VTT or text
Chat shortcut: /subtitles
Caption generator without typing: upload video or audio, choose a language or have it detected automatically, download the SRT or VTT file and proofread it. It can be loaded into YouTube Studio and common editing programs.
With Account
Cost: 1,400 tokens per minute or part thereof · Input and result are automatically deleted 24 hours after upload.
This tool can be used with a free account. Sign up for free
From the spoken word to the finished subtitle file
A large share of videos in social media feeds plays without sound. Subtitles make sure the message still gets through, make content accessible to people with hearing impairments and give search engines text they can read. Done by hand, typing and timing takes several times as long as the video itself.
The tool extracts the audio track from your file, converts it to compact mono audio, runs speech recognition to determine the text with timings and outputs it in the blocks the recognition sets; no custom rules for line length or reading speed are applied. You get the result as SRT with number, time span and text, as VTT for web players or as a plain transcript. You specify the language or have it detected.
Typical cases: subtitling YouTube videos and tutorials, securing interviews and podcasts as text, making lectures and meetings readable, or giving reels a text track before burning it in, which you proofread first.
How it works
- 1
Upload file
Drop in or choose video or audio up to 96 MB and 50 minutes; split longer recordings first in an editing program, audio up to 60 minutes and 50 MB also with “MP3 Cutter”.
- 2
Choose language
“Detect automatically” or set one of the ten languages explicitly if the recording contains dialect or many names.
- 3
Choose format
SRT for YouTube and editing programs, VTT for web video or “Text only” for a transcript without timings.
- 4
Download and proofread
After “Let's go”, check the text, save the file and load it into YouTube Studio or your editing program.
Typical uses
YouTube and tutorials
Upload the SRT so viewers can follow with and without sound and the video is easier to find.
Reels, Shorts, TikTok
Generate the text track, check it and then burn it in in the editing program or app.
Podcasts and interviews
Transcript without timings as the basis for show notes, quotes or blog posts.
Lectures and meetings
Store the recording as VTT or text so participants can read and search it.
Good to know
- Input: MP4, MOV, WebM, MKV, MP3, M4A, WAV, OGG, FLAC up to 96 MB and 50 minutes per file. Output: SRT, VTT or plain text.
- Languages: detect automatically, or German, English, French, Italian, Spanish, Portuguese, Dutch, Polish, Turkish, Russian.
- The audio track is reduced to mono MP3 at 16 kHz and 64 kbit/s before recognition; timestamps and blocks come from the recognition as they are, so proofread proper names, technical terms, music passages and timings.
- Cost: fixed token amount per minute or part thereof, shown after selecting the file; account required, daily allowance, packs and subscriptions.
- The audio track goes to the AI provider for recognition; upload and result are deleted on our side after 24 hours (hourly cleanup).
- Chat shortcut: /subtitles. Burning into the video and translating the subtitles are separate steps in your editing program.
Frequently asked questions
How accurate are the subtitles?
Usually good with clear speech without music. You should proofread names, technical terms, strong dialects, passages with several speakers at once and the timings, and adjust them in the editing program.
SRT, VTT or text only: what do I need?
SRT is accepted by YouTube, editing programs and most players. VTT is the format for videos on websites. “Text only” delivers the plain transcript without timings, for example for descriptions or blog posts.
How do I get the subtitles into the video?
Upload the SRT file in YouTube Studio under “Subtitles” or import it in the editing program, where you can adjust lines. Burning them permanently into the video is also done there or in the app you publish the video with.
Why are accented characters such as é, ñ or ü displayed wrong?
The file comes as UTF-8, which YouTube and current programs read correctly. If an older player shows garbled characters, switch its encoding to UTF-8.
Which languages are recognized?
The available languages are German, English, French, Italian, Spanish, Portuguese, Dutch, Polish, Turkish and Russian. With “Detect automatically” the AI determines the language itself; with mixed recordings, setting it explicitly helps.
Which files work and what does it cost?
MP4, MOV, WebM, MKV, MP3, M4A, WAV, OGG and FLAC up to 96 MB and 50 minutes; split longer recordings first in an editing program, audio up to 60 minutes and 50 MB also with “MP3 Cutter”. A fixed token amount is charged per minute or part thereof; you see the exact number after selecting the file. The tool requires an account.
What happens to my file?
The audio track is reduced and passed to the AI provider for recognition; what it stores is governed by its terms. On our side, upload and result are deleted 24 hours after upload (hourly cleanup).