In this short guide, you'll see how to automatically generate YouTube subtitles using Python and OpenAI Whisper.

Here you can find the short answer:

(1) Using OpenAI Whisper

import whisper
model = whisper.load_model("base")
result = model.transcribe("video.mp4")

(2) Generate SRT files with timestamps

from whisper.utils import get_writer
writer = get_writer("srt", "output/")
writer(result, "video.srt")

So let's see several methods to generate accurate subtitles for YouTube videos automatically.

1: Install and setup OpenAI Whisper

Let's start with installing OpenAI Whisper, the most accurate speech-to-text model for subtitle generation:

pip install openai-whisper

Basic transcription:

import whisper

model = whisper.load_model("base")

audio_file = "video.mp4"
result = model.transcribe(audio_file)

print("Transcription:")
print(result["text"])

result will be:

Transcription:
Welcome to this tutorial on Python programming. Today we'll learn about automatic subtitle generation using OpenAI Whisper. This powerful tool can transcribe audio with high accuracy in multiple languages.

Available Whisper models (by size and accuracy):

Model Size Speed Accuracy
tiny 39 MB Fastest Good
base 74 MB Fast Better
small 244 MB Medium Great
medium 769 MB Slow Excellent
large 1550 MB Slowest Best

Choose "base" for quick testing, "small" for production balance, "large" for maximum accuracy.

2: Generate SRT subtitle files with timestamps

The most useful output format for YouTube is SRT (SubRip) which includes timestamps for each subtitle segment:

import whisper
from whisper.utils import get_writer

model = whisper.load_model("base")

audio_file = "tutorial_video.mp4"
result = model.transcribe(audio_file)

output_dir = "subtitles/"
srt_writer = get_writer("srt", output_dir)

srt_writer(result, audio_file)

print(f"SRT file created: {output_dir}tutorial_video.srt")

result:

SRT file created: subtitles/tutorial_video.srt

Generated SRT file content:

1
00:00:00,000 --> 00:00:03,500
Welcome to this tutorial on Python programming.

2
00:00:03,500 --> 00:00:07,200
Today we'll learn about automatic subtitle generation.

3
00:00:07,200 --> 00:00:11,800
This powerful tool can transcribe audio with high accuracy.

This SRT file can be directly uploaded to YouTube, added to video editors, or used with media players.

3: Generate multilingual subtitles

Whisper supports 99 languages automatically. Specify the language for better accuracy or auto-detect:

import whisper

model = whisper.load_model("base")

audio_file = "spanish_video.mp4"

result_auto = model.transcribe(audio_file)
result_spanish = model.transcribe(audio_file, language="es")

print(f"Auto-detected language: {result_auto['language']}")
print(f"\nTranscription: {result_spanish['text'][:100]}...")

result:

Auto-detected language: es

Transcription: Bienvenidos a este tutorial sobre programación en Python. Hoy aprenderemos sobre la generación...

Popular language codes:

  • English: en
  • Spanish: es
  • French: fr
  • German: de
  • Japanese: ja
  • Chinese: zh
  • Hindi: hi

4: Batch process multiple videos

For YouTube channels or course creators, process multiple videos efficiently:

import whisper
import os
from pathlib import Path

video_folder = "videos/"
output_folder = "subtitles/"

model = whisper.load_model("small")

video_files = list(Path(video_folder).glob("*.mp4"))

print(f"Processing {len(video_files)} videos...")

for i, video_path in enumerate(video_files, 1):
    print(f"\n[{i}/{len(video_files)}] Processing: {video_path.name}")
    
    result = model.transcribe(str(video_path))
    
    from whisper.utils import get_writer
    srt_writer = get_writer("srt", output_folder)
    srt_writer(result, video_path.stem)
    
    print(f"✓ Created: {output_folder}{video_path.stem}.srt")

print(f"\nCompleted! {len(video_files)} subtitle files generated.")

result:

Processing 3 videos...

[1/3] Processing: lecture_01.mp4
✓ Created: subtitles/lecture_01.srt

[2/3] Processing: lecture_02.mp4
✓ Created: subtitles/lecture_02.srt

[3/3] Processing: lecture_03.mp4
✓ Created: subtitles/lecture_03.srt

Completed! 3 subtitle files generated.

This saves hours of manual transcription for video series or course content.

Free Alternative Tools

For Windows users without Python:

  1. OpenAI Whisper Desktop - GUI wrapper for Whisper

    • Download: GitHub
    • Drag-and-drop interface
    • No coding required

  2. YouTube Subtitles Editor

    • Built into YouTube Studio
    • Auto-generates subtitles
    • Manual editing interface
    • Free for all creators

  3. Subtitle Edit

    • Open-source subtitle editor
    • Auto-transcription integration
    • Works with Whisper
    • Download: subtitleedit.org

Resources