ai-video-generator

AI Video Generator — Short-Form Content Pipeline

Overview

Automate creation of short-form videos (TikTok, YouTube Shorts, Instagram Reels) using AI for every step: topic research, script writing, text-to-speech narration, stock footage matching, subtitle generation, and final assembly. Inspired by MoneyPrinterTurbo (53k+ stars).

Instructions

Step 1: Set Up the Environment

pip install anthropic openai requests moviepy pydub whisperx srt
sudo apt install ffmpeg  # Linux — or: brew install ffmpeg (macOS)

API keys needed: Anthropic or OpenAI (scripts), ElevenLabs or OpenAI TTS (voice), Pexels (free stock footage).

Step 2: AI Script Writing

import anthropic

def generate_script(topic, duration_seconds=45):
    """Generate a video script optimized for short-form content."""
    client = anthropic.Anthropic()
    prompt = f"""Write a {duration_seconds}-second video script about: {topic}
    Format:
    HOOK (first 3 seconds): A shocking statement or question that stops scrolling
    BODY (main content): 3-5 punchy facts or points, each 1-2 sentences
    CTA (last 5 seconds): Call to action — follow, like, comment
    Rules:
    - Conversational, no complex sentences
    - Each sentence on its own line
    - ~{duration_seconds * 2.5:.0f} words ({duration_seconds}s at 150wpm)
    - Use power words: secret, shocking, nobody tells you, actually
    - No emojis or hashtags — this is a voiceover script
    """
    response = client.messages.create(
        model="claude-sonnet-4-20250514", max_tokens=500,
        messages=[{"role": "user", "content": prompt}]
    )
    return response.content[0].text

Step 3: Text-to-Speech Narration

import requests, os

def generate_voice_elevenlabs(text, output_path='narration.mp3'):
    """Generate voiceover using ElevenLabs."""
    url = "https://api.elevenlabs.io/v1/text-to-speech/21m00Tcm4TlvDq8ikWAM"
    headers = {"xi-api-key": os.environ["ELEVENLABS_API_KEY"], "Content-Type": "application/json"}
    data = {"text": text, "model_id": "eleven_turbo_v2_5",
            "voice_settings": {"stability": 0.5, "similarity_boost": 0.75}}
    response = requests.post(url, json=data, headers=headers)
    with open(output_path, 'wb') as f:
        f.write(response.content)
    return output_path

def generate_voice_openai(text, output_path='narration.mp3'):
    """Generate voiceover using OpenAI TTS (cheaper alternative)."""
    from openai import OpenAI
    client = OpenAI()
    response = client.audio.speech.create(model="tts-1-hd", voice="onyx", input=text)
    response.stream_to_file(output_path)
    return output_path

Step 4: Stock Footage Selection

def search_pexels_videos(query, count=5):
    """Search Pexels for portrait-oriented stock video clips."""
    url = "https://api.pexels.com/videos/search"
    headers = {"Authorization": os.environ["PEXELS_API_KEY"]}
    params = {"query": query, "per_page": count, "orientation": "portrait", "size": "medium"}
    response = requests.get(url, headers=headers, params=params)
    videos = response.json().get('videos', [])
    results = []
    for v in videos:
        files = sorted(v['video_files'], key=lambda x: x.get('height', 0), reverse=True)
        hd = next((f for f in files if f.get('height', 0) >= 720), files[0])
        results.append({'id': v['id'], 'url': hd['link'], 'duration': v['duration']})
    return results

Step 5: Subtitle Generation

def generate_subtitles(audio_path, output_srt='subtitles.srt'):
    """Generate word-level subtitles using WhisperX."""
    import whisperx, srt
    from datetime import timedelta
    model = whisperx.load_model("base", device="cpu")
    audio = whisperx.load_audio(audio_path)
    result = model.transcribe(audio)
    align_model, metadata = whisperx.load_align_model(language_code="en")
    aligned = whisperx.align(result["segments"], align_model, metadata, audio)
    subs = []
    words = [w for seg in aligned["segments"] for w in seg.get("words", [])]
    for i in range(0, len(words), 4):
        group = words[i:i + 4]
        if not group: continue
        start = timedelta(seconds=group[0].get('start', 0))
        end = timedelta(seconds=group[-1].get('end', 0))
        text = ' '.join(w['word'] for w in group)
        subs.append(srt.Subtitle(index=len(subs)+1, start=start, end=end, content=text))
    with open(output_srt, 'w') as f:
        f.write(srt.compose(subs))
    return output_srt

Step 6: Video Assembly with FFmpeg

import subprocess

def assemble_video(clips, narration, subtitles, output='final.mp4'):
    """Assemble final video: concatenate clips, add narration and subtitles."""
    concat_list = 'concat_list.txt'
    with open(concat_list, 'w') as f:
        for clip in clips:
            f.write(f"file '{clip}'\n")
    subprocess.run([
        'ffmpeg', '-y', '-f', 'concat', '-safe', '0', '-i', concat_list,
        '-vf', 'scale=1080:1920:force_original_aspect_ratio=increase,crop=1080:1920',
        '-c:v', 'libx264', '-preset', 'fast', '-an', 'temp_video.mp4'
    ], check=True)
    subtitle_filter = (f"subtitles={subtitles}:force_style='"
        "FontName=Arial,FontSize=18,PrimaryColour=&H00FFFFFF,"
        "OutlineColour=&H00000000,Outline=2,Bold=1,Alignment=2'")
    subprocess.run([
        'ffmpeg', '-y', '-i', 'temp_video.mp4', '-i', narration,
        '-vf', subtitle_filter, '-c:v', 'libx264', '-c:a', 'aac',
        '-shortest', output
    ], check=True)
    return output

Step 7: Full Pipeline

def generate_video(topic, output_dir='./output'):
    """Complete pipeline: topic -> finished video."""
    import os
    os.makedirs(output_dir, exist_ok=True)
    script = generate_script(topic)
    narration = generate_voice_elevenlabs(script, f'{output_dir}/narration.mp3')
    keywords = topic.split()[:3]
    videos = search_pexels_videos(' '.join(keywords), count=3)
    clips = []
    for i, v in enumerate(videos):
        path = f'{output_dir}/clip_{i}.mp4'
        requests.get(v['url'], stream=True)  # download clip
        clips.append(path)
    subs = generate_subtitles(narration, f'{output_dir}/subs.srt')
    return assemble_video(clips, narration, subs, f'{output_dir}/final.mp4')

Examples

Example 1: Generate a Batch of Tech Fact Videos

A creator produces 5 technology-themed short videos for TikTok in one run:

topics = [
    "AI tools nobody talks about",
    "Apps that feel illegal to use for free",
    "Websites that will blow your mind",
    "Free AI tools every student needs",
    "Tech gadgets under $50 that changed my life"
]
for topic in topics:
    output = generate_video(topic, output_dir=f'./output/{topic[:30]}')
    print(f"Video ready: {output}")
# Each video: ~45 seconds, portrait 1080x1920, with subtitles and narration
# Total cost: ~$0.25 (5 x $0.05 per video) using ElevenLabs + Claude Sonnet

Example 2: Daily Finance Shorts for YouTube

A finance channel automates daily Shorts upload with trending money topics:

import schedule

def daily_finance_video():
    topic = "3 passive income ideas that actually work in 2025"
    script = generate_script(topic, duration_seconds=55)
    # Script output:
    # HOOK: "You're losing money every single day you don't know about these."
    # BODY: 1. Print-on-demand stores ($500-2k/mo)
    #       2. AI-generated content licensing ($300-1k/mo)
    #       3. Dividend ETF stacking ($200-800/mo passive)
    # CTA: "Follow for more money tips that nobody tells you about."
    narration = generate_voice_openai(script, './daily/narration.mp3')
    videos = search_pexels_videos("money finance investing", count=4)
    # Downloads 4 portrait clips of money/charts/lifestyle footage
    # Assembles with bold white subtitles, outputs 55-second Short
    final = assemble_video(['./daily/clip_0.mp4', './daily/clip_1.mp4'],
                           narration, './daily/subs.srt', './daily/final.mp4')

schedule.every().day.at("08:00").do(daily_finance_video)

Guidelines

• Respect copyright — only use royalty-free stock footage (Pexels, Pixabay) or your own content
• Disclose AI usage — YouTube and TikTok require disclosure of AI-generated content
• Review before publishing — always watch the final video; AI scripts can contain inaccuracies
• Optimize for the first 3 seconds — the hook determines whether viewers stay or scroll
• Test multiple voices — ElevenLabs offers dozens of voices; find one that fits your niche
• Monitor performance — track views, retention, and click-through to iterate on content style

References

• MoneyPrinterTurbo — original inspiration (53k stars)
• Pexels API — free stock video
• ElevenLabs — realistic text-to-speech
• WhisperX — word-level subtitle alignment