How to Download YouTube Transcript as SRT File [2026]

1
00:00:00,000 --> 00:00:03,500
Welcome to this video about
YouTube transcripts.

2
00:00:03,500 --> 00:00:07,200
Today we'll cover everything
you need to know.

3
00:00:07,200 --> 00:00:11,000
Let's start with the basics
of transcript extraction.

HH:MM:SS,mmm --> HH:MM:SS,mmm

HH = Hours (00-99)
MM = Minutes (00-59)
SS = Seconds (00-59)
mmm = Milliseconds (000-999)

# Windows (with pip)
pip install yt-dlp

# Mac (with Homebrew)
brew install yt-dlp

# Linux
pip install yt-dlp

# Download auto-generated subtitles as SRT
yt-dlp --write-auto-sub --sub-format srt --skip-download "VIDEO_URL"

# Download manual subtitles if available
yt-dlp --write-sub --sub-format srt --skip-download "VIDEO_URL"

# Download both auto and manual if available
yt-dlp --write-auto-sub --write-sub --sub-format srt --skip-download "VIDEO_URL"

# English only
yt-dlp --write-auto-sub --sub-lang en --sub-format srt --skip-download "VIDEO_URL"

# Multiple languages
yt-dlp --write-auto-sub --sub-lang en,es,fr --sub-format srt --skip-download "VIDEO_URL"

ffmpeg -i subtitles.vtt subtitles.srt

import webvtt

vtt = webvtt.read('subtitles.vtt')
srt_content = ""

for i, caption in enumerate(vtt, 1):
    start = caption.start.replace('.', ',')
    end = caption.end.replace('.', ',')
    srt_content += f"{i}\n{start} --> {end}\n{caption.text}\n\n"

with open('subtitles.srt', 'w') as f:
    f.write(srt_content)

GOOD:
1
00:00:00,000 --> 00:00:03,000
This is a good caption length
for easy reading.

BAD:
1
00:00:00,000 --> 00:00:03,000
This is way too long of a caption that will be difficult to read quickly while also watching the video content.

WRONG:
1
00:00:00,000 --> 00:00:03,500
First caption

2
00:00:03,000 --> 00:00:06,000
Second caption (overlaps!)

CORRECT:
1
00:00:00,000 --> 00:00:03,000
First caption

2
00:00:03,100 --> 00:00:06,000
Second caption

# Shift all captions +2 seconds
00:00:05,000 --> 00:00:08,000
becomes
00:00:07,000 --> 00:00:10,000

# Correct errors
"elan musk" → "Elon Musk"

# Before
1
00:00:00,000 --> 00:00:06,000
This is a very long caption that should be split into two separate captions for better readability.

# After
1
00:00:00,000 --> 00:00:03,000
This is a very long caption that
should be split into two

2
00:00:03,000 --> 00:00:06,000
separate captions for
better readability.