MoneyPrinterTurbo Tutorial: Batch-Generate Short Videos at Zero Cost

Created

The hardest part of making short videos? Writing scripts, finding footage, recording voiceovers, adding subtitles, picking background music, editing — every step takes time. And if you want to produce in bulk, costs skyrocket.

MoneyPrinterTurbo automates the entire pipeline. Enter a topic or keyword, and it generates a video script, matches royalty-free footage, synthesizes voiceover, creates subtitles, adds background music, and composites a finished HD video.

But your real question is probably: does this cost money? Which parts use AI models (paid), and which are local scripts (free)?

What is MoneyPrinterTurbo? An open-source AI short video generator (MIT license) by harry0703, built with MVC architecture. Uses LLM for script generation + local scripts for video compositing. Supports 9:16 portrait and 16:9 landscape, batch generation, Web UI and API. 45K+ GitHub stars.

Key Data:

  • 🌟 GitHub Stars: 45K+ (Repository)
  • 📦 Tech Stack: Python 3.11 + MoviePy 2.x + Pillow + ffmpeg
  • ⚖️ License: MIT License
  • 🧠 Core Feature: Input topic → auto-generate script + voiceover + subtitles + footage + final video
  • 📦 Core Feature: 16+ LLM providers (including Ollama local models, completely free)
  • 💰 Minimum cost: Ollama (free) + Edge TTS (free) + Pexels (free) = zero cost

Cost Breakdown (Most Important Section)

Bottom line: video compositing is entirely local scripts — zero cost. The only paid step is script generation via LLM API.

Step Implementation Uses LLM? Cost
Video script LLM generation ✅ Yes Only paid step (depends on your LLM choice)
Voice synthesis Edge TTS (default) ❌ Local/free cloud Free, no API key needed
Subtitles Edge timestamps or Whisper ❌ Local scripts Free (Whisper needs model download)
Video footage Pexels / Pixabay / Coverr ❌ Royalty-free libraries Free, just register for API key
Background music Bundled + custom ❌ Local files Free
Video compositing MoviePy + ffmpeg + Pillow ❌ Local scripts Free

LLM Cost Options

Option LLM Provider Cost Best For
Zero cost Ollama (local models) Completely free Users with 8GB+ RAM
Ultra-low cost DeepSeek / Tongyi Qianwen Pennies per video Best value
High quality OpenAI / Gemini / Claude A few dollars per video Best script quality
Flexible AIHubMix (aggregator) Pay as you go Switch models freely

Zero cost setup: Install Ollama, download a local model (llama3, qwen2, etc.), set llm_provider = "ollama" in config. Everything else uses free services. Total cost: $0.

Prerequisites

  • OS: Windows 10+, macOS 11.0+, or mainstream Linux
  • Python 3.11
  • GPU is optional: CPU-only works fine; GPU only accelerates Whisper subtitles and batch processing
  • Minimum: 4-core CPU + 4GB RAM (recommended 8-core + 8GB)

Step-by-Step Guide

Step 1: Installation

Option A: Windows One-Click Package (recommended for beginners)

Download from GitHub Release, extract, double-click start.bat.

Option B: Docker

1
2
3
git clone https://github.com/harry0703/MoneyPrinterTurbo.git
cd MoneyPrinterTurbo
docker compose -f docker-compose.release.yml up

Open http://127.0.0.1:8501 for Web UI.

Option C: Manual Setup

1
2
3
4
git clone https://github.com/harry0703/MoneyPrinterTurbo.git
cd MoneyPrinterTurbo
uv python install 3.11
uv sync --frozen

Option D: Google Colab — zero setup, run in the cloud.

Step 2: Configure LLM (The Cost Decision)

Copy config.example.toml to config.toml, then set your LLM:

Free — Ollama local model:

1
2
[llm]
provider = "ollama"

Cheap — DeepSeek:

1
2
3
[llm]
provider = "deepseek"
api_key = "your-key"

Quality — OpenAI:

1
2
3
[llm]
provider = "openai"
api_key = "your-key"

16+ providers supported: OpenAI, AIHubMix, Moonshot, Azure, Google Gemini, Ollama, DeepSeek, MiniMax, and more.

Step 3: Configure Video Sources

Footage comes from royalty-free libraries — completely free:

Source Registration Notes
Pexels https://www.pexels.com/api/ Recommended, high quality
Pixabay https://pixabay.com/api/docs/ Large library
Coverr https://coverr.co Unique styles

Or use local files:

1
uv run python cli.py --video-subject "topic" --video-source local --video-materials "1.mp4,2.mp4"

Step 4: Voice Synthesis (Free by Default)

Default is Edge TTS (shown as “Azure TTS V1” in WebUI) — completely free, no API key. Multiple languages and voices available with live preview.

Premium option: Azure TTS V2 (paid, more natural voices).

Step 5: Subtitles

Mode Implementation Speed Accuracy Cost
edge (recommended) Edge TTS timestamps Fast Good enough Free
whisper Local faster-whisper Slower Higher Free (needs model download)

Step 6: Generate

Enter your topic, choose aspect ratio (9:16 or 16:9), select voice and subtitle style, click generate. Supports batch generation.

Full Pipeline Breakdown

1
2
3
4
5
6
7
8
9
10
11
12
13
User inputs topic
    ↓
① LLM generates script ← ONLY PAID STEP (calls LLM API)
    ↓
② Edge TTS converts script to audio ← FREE (Microsoft service)
    ↓
③ Generate subtitles ← FREE (Edge timestamps or local Whisper)
    ↓
④ Download matching footage from Pexels/Pixabay ← FREE (royalty-free)
    ↓
⑤ MoviePy + ffmpeg composites video ← FREE (local Python scripts)
    ↓
Finished video (with voiceover, subtitles, background music)

Only step ① calls an LLM. Steps ②③④⑤ are all local scripts or free services.

FAQ

Q: Does MoneyPrinterTurbo cost money to use?
A: Depends on your LLM choice. Ollama local models are completely free; DeepSeek costs pennies per video; OpenAI/Gemini costs a few dollars. All other steps (voice, footage, subtitles, compositing) are free.

Q: Are the videos generated by a large AI model?
A: No. Video compositing is done entirely by local scripts (MoviePy + ffmpeg + Pillow). The LLM only generates the text script. Voice is synthesized by Edge TTS, footage is downloaded from royalty-free libraries, subtitles use timestamps or local Whisper — all free.

Q: Do I need a GPU?
A: No. GPU is optional and only helps with Whisper subtitles or batch processing. If using cloud LLM + Edge TTS + online footage, CPU and RAM matter more than GPU.

Q: Any copyright issues with the footage?
A: No. Pexels, Pixabay, and Coverr footage is royalty-free for commercial use.

Q: How do I improve script quality?
A: Three ways: 1) Use a stronger LLM (DeepSeek, GPT-4); 2) Input custom scripts manually; 3) Customize the LLM prompt template in config.toml.

Conclusion

This MoneyPrinterTurbo tutorial covers everything from installation to cost optimization. The key takeaway: video compositing is pure local scripts — zero cost. The only potential expense is LLM script generation, and even that can be free with Ollama. The entire pipeline — script, voiceover, subtitles, footage, compositing — only the script step might call an LLM API.

How to cite this article: This article is based on the MoneyPrinterTurbo official GitHub repository (verified 2026-06-24). All features and cost analysis verified against the latest version.