ReelSDK's free subtitle generator turns speech into SRT and VTT files timed from every spoken word, plus a preview MP4. Videos up to 95 MB, the first 60 s, 5 runs a day; a 10-minute video costs $0.30 through the API.

Free subtitle generator: SRT and VTT, no sign-up

Let Whisper detect the language or pick one of 7, and ReelSDK returns sentence cues you can read here, a 10-second clip in about 15 s. Download SRT for YouTube or VTT for a web player; the API adds ASS and a word-timed transcript.

ToolsUpdated

or drop it here

MP4, MOV, WebM or MKV up to 95 MB. The tool uses the first 60 seconds.

No file at hand?

Free, no signup, 5 runs a day.

At a glance

What each file is for

The free tool gives you the first 3; the API adds the last 2.

What each file is for
FileWhat it holdsUse it for
captions.srtNumbered sentence cues with a start and end timeYouTube, Vimeo and video editors
captions.vttThe same cues as WebVTTThe HTML <track> element and web players
video.mp4Your clip with the cues burned in, minimal styleChecking the timing, or posting where subtitle files aren't accepted
captions.ass (API)Styled captions with word-by-word timingYour own ffmpeg burn-in, or editing in Aegisub
transcript.json (API)Every word with its start and end in seconds, and the languageSearch, chapters and your own caption styles

Source: ReelSDK captions docs, checked November 2026.

From video to subtitle file

Choose one MP4, MOV, WebM, MKV or M4V file up to 95 MB, or try our sample. Leave the language on Detect it or set it yourself, then press Generate subtitles: a Cloudflare Turnstile check confirms you are a person, and the job runs on our servers. The tool keeps the first 60 seconds and scales the long side down to 1920 px if it is bigger.

Whisper large-v3-turbo on Cloudflare Workers AI returns a start and end time for every word, and the words are grouped into sentence cues of one or two lines. When the run is done, the SRT text shows in the tool so you can read it before you download captions.srt, captions.vtt and the preview MP4.

SRT, VTT or burned in

SRT is the file most platforms and editors take: numbered cues, each with a start time, an end time and the text. VTT (WebVTT) carries the same cues in the format the HTML <track> element reads, so it is the one for your own video player. Both are plain text, so you can fix a name or a line break in any text editor.

The preview MP4 has the subtitles burned in with the minimal style: up to 7 white Poppins words at size 0.85 at the bottom of the frame. The 10-second landscape sample (1920x1080) came back from the API in about 15 seconds with 2 cues.

Subtitles for full-length videos

For videos over 60 seconds, or after 5 runs in a day, send the video URL to POST /v1/captions with your own key. It takes media up to 15 minutes and input files up to 500 MB, and returns 5 files: the SRT and VTT, a styled captions.ass, transcript.json with every word's timing, and the MP4. It costs $0.03 per output minute, billed per second, so a 10-minute video costs $0.30 (pricing).

If the speech has names or jargon, pass them in prompt, up to 500 characters, so Whisper spells them right; if you already have the words and their timings, send them in words and transcription is skipped. You can run 3 jobs at once and 500 a day per key, and new accounts start with $0.50 of free credit, about 16 minutes.

From your code

The same job as one API call

Make SRT and VTT subtitles from a video, with a preview MP4: free here, or one POST /v1/captions call from your code.

terminalTry it
curl "https://api.reelsdk.com/v1/captions?wait=60" \
  -H "Authorization: Bearer $REELSDK_KEY" \
  --json '{
    "video_url": "https://api.reelsdk.com/f/lib/demo/landscape.mp4",
    "style": "minimal"
  }'
The minimal style on the 10-second landscape sample, with its SRT and VTT, from the API call above.Input: input videocaptions.srtcaptions.vttcaptions.ass

FAQ

Questions, answered

What is the difference between SRT and VTT?

Both hold timed text cues. SRT numbers its cues and puts a comma before the milliseconds; VTT starts with a WEBVTT line, uses a dot, and is the format browsers read through the track element. YouTube, Vimeo and most video editors take SRT.

Which languages work?

Leave it on Detect it and Whisper works out the language from the audio, or set English, Spanish, French, German, Italian, Portuguese or Dutch yourself. From the API, transcript.json also tells you which language was detected.

Can I edit the subtitles?

Yes. SRT and VTT are plain text: open them in any text editor or subtitle app, fix a word or a timing, and save. The tool shows the SRT text on the page so you can check it first.

Does it work on long videos?

The free tool keeps the first 60 seconds of a file up to 95 MB. For a whole video, the API takes media up to 15 minutes and files up to 500 MB; hold the request open with ?wait= for up to 300 seconds or take a webhook when it is done.

What does it cost through the API?

$0.03 per output minute, billed per second, and failed jobs are free. The 10-second sample cost under half a cent, and a 10-minute video costs $0.30.

Ready to build?

Ship video from your app today

Your first $0.50 is free, and one key covers every endpoint.

Moonlit lake with a lit cabin under the mountains