One POST /v1/captions to ReelSDK turns a video into SRT and VTT subtitles, a word-timed transcript and a captioned copy at $0.03 per output minute, so a 60-second clip costs about 3 cents.

Turn a video into SRT and VTT subtitles with one API call

Send one video URL and in under 30 seconds ReelSDK sends back 5 files for a 1-minute video, the SRT and VTT among them, with every cue timed to the speech.

Use casesUpdated

A 60-second NASA ScienceCasts clip in; SRT, VTT, transcript and a captioned copy out, in under 30 seconds.Input: source clipcaptions.srtcaptions.vtttranscript.json

The request

POST /v1/captions

Send a video URL and get SRT and VTT subtitle files, a word-timed transcript and a captioned copy of the video back.

terminalTry it
curl "https://api.reelsdk.com/v1/captions?wait=300" \
  -H "Authorization: Bearer $REELSDK_KEY" \
  --json '{
    "video_url": "https://api.reelsdk.com/f/lib/in/nasa-sciencecasts/eclipse-60s.mp4",
    "style": "minimal",
    "language": "en",
    "prompt": "NASA, ScienceCasts, annular eclipse"
  }'
request.mjsTry it
const res = await fetch("https://api.reelsdk.com/v1/captions?wait=300", {
  method: "POST",
  headers: { Authorization: `Bearer ${process.env.REELSDK_KEY}`, "Content-Type": "application/json" },
  body: JSON.stringify({
    "video_url": "https://api.reelsdk.com/f/lib/in/nasa-sciencecasts/eclipse-60s.mp4",
    "style": "minimal",
    "language": "en",
    "prompt": "NASA, ScienceCasts, annular eclipse"
  }),
});
const { job } = await res.json();
response.json
{
  "job": {
    "id": "job_jz0g8m9ogmtvnafc399p",
    "op": "captions",
    "status": "succeeded",
    "stage": null,
    "created_at": "2026-11-07T01:31:37.398Z",
    "finished_at": "2026-11-07T01:32:01.406Z",
    "outputs": {
      "video": "https://api.reelsdk.com/f/out/job_jz0g8m9ogmtvnafc399p/video.mp4?s=...",
      "transcript": "https://api.reelsdk.com/f/out/job_jz0g8m9ogmtvnafc399p/transcript.json?s=...",
      "srt": "https://api.reelsdk.com/f/out/job_jz0g8m9ogmtvnafc399p/captions.srt?s=...",
      "vtt": "https://api.reelsdk.com/f/out/job_jz0g8m9ogmtvnafc399p/captions.vtt?s=...",
      "ass": "https://api.reelsdk.com/f/out/job_jz0g8m9ogmtvnafc399p/captions.ass?s=..."
    },
    "result": null,
    "duration_s": 60.027,
    "error": null,
    "usage": {
      "output_seconds": 60.027,
      "audio_seconds": 60,
      "tts_characters": 0,
      "images": 0
    },
    "poll_url": "https://api.reelsdk.com/v1/jobs/job_jz0g8m9ogmtvnafc399p"
  }
}

What comes back

One request returns 5 files. captions.srt and captions.vtt hold the subtitle cues, each with a start and an end time. transcript.json has the full text, the detected language and every word with its own start and end time in seconds, so you can build your own cue layout.

The job also returns captions.ass and video.mp4, a 1280x720 copy of the clip with the captions burned in. This run used the minimal style, which puts Poppins at the bottom at size 0.85, 7 words per group. You can see the first three cues from this clip next to this text.

captions.srt
1
00:00:00,000 --> 00:00:01,680
Mark your calendar.

2
00:00:01,680 --> 00:00:06,020
On Sunday, May 20th, the Sun is
going to turn into a ring of fire.

3
00:00:06,020 --> 00:00:11,640
It's an annular solar eclipse, the
first one in the USA in almost 18 years.

At a glance

The 5 files one request returned

The 5 files one request returned
FileSizeWhat it is for
captions.srt1.6 KBSubRip cues for YouTube uploads, video players and editors
captions.vtt1.6 KBWebVTT cues for an HTML5 <track> element
transcript.json8.4 KBThe full text, the detected language, the duration and every word with its start and end in seconds
captions.ass49 KBThe styled captions as burned in, for editors that read ASS
video.mp45.2 MBThe video at 1280x720 with the captions burned in

Sources: Sample run, ReelSDK captions docs, checked November 2026.

Getting names and terms right

The speech is transcribed with whisper-large-v3-turbo, with whisper as the fallback. The prompt field takes up to 500 characters of names and terms so the transcriber spells them right. For this clip the request sent NASA, ScienceCasts, annular eclipse.

Pass language to set the spoken language, or leave it out and it is detected. If you already have words with timings from your own transcriber or a script, pass them in words (up to 50,000) and the transcription step is skipped.

Price and limits

Captions cost $0.03 per output minute with transcription included. This clip is a minute of output and was done in 24 seconds, so the job came to about 3 cents, and the $0.50 signup credit covers 2 clips like this one, because starting a job holds the most it can cost, about 45 cents, until it finishes.

Input files can be up to 500 MB, media can be up to 15 minutes long, and a job can run for up to 10 minutes. The output links stay up for 7 days. You can check the full field list in the captions docs, see styled burned-in captions on captions, and check prices on pricing.

What it costs

What this clip cost

Video, 60 s at $0.03 per minute
$0.03
Total
$0.03

Billed per second of output, transcription included. A failed job costs nothing.

Limits

What one request takes

Word list
Up to 50,000 words
Spelling hint
prompt up to 500 characters
Input files
500 MB each, 1 GB per job
Media length
Up to 15 minutes for captions and slideshows
Job run time
Up to 10 minutes
Concurrent jobs
3 per key
Jobs per day
500 per key
Output retention
7 days

How it works

From request to file

  1. Host the video

    Put the video at a public https URL, or upload it first with POST /v1/uploads. Inputs can be up to 500 MB and 15 minutes long.

  2. Send POST /v1/captions

    Call POST /v1/captions with video_url, a style, and language and a prompt with names if you know them.

  3. Wait or get a webhook

    Add ?wait=300 to get the finished job back on the same request, or pass a webhook_url.

  4. Save the files

    Download captions.srt and captions.vtt from the output links before they expire after 7 days.

FAQ

Questions, answered

What does a subtitle file cost?

Captions cost $0.03 per output minute, with transcription included. This 1-minute clip cost about 3 cents, so the $0.50 signup credit covers 2 clips like it, because starting a job holds about 45 cents until it finishes.

Can I get only the SRT without a burned video?

Every captions job returns the captioned video too. Just grab the SRT and VTT files and ignore the MP4, since the price is the same.

Which languages does it transcribe?

Set language to the spoken language, or leave it out to have it detected; the detected language is saved in transcript.json. Japanese, Chinese and Korean captions are drawn in Noto Sans.

How do I fix a misspelled name?

Pass the names and terms in the prompt field, up to 500 characters, so the transcriber spells them right.

How long does it take?

Under 30 seconds for this 1-minute clip, from the request to the finished job. Add ?wait=300 to get it back on the same request.

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