I just generated an entire short-form video from scratch using MoneyPrinterTurbo. Script, footage, voiceover, subtitles, background music. All free. Here's exactly how to do it.

Step 1: Get the Tool

Go to the MoneyPrinterTurbo GitHub repository. It's open source, MIT license, and completely free.

Step 2: Open It in Google Colab

The easiest way to use this is Google Colab — no installation required. You'll find the Colab link directly in the repo. Click it and it opens a ready-to-run notebook in your browser.

Step 3: Connect to GPU

In the first cell, change your runtime to T4 GPU. This makes everything run significantly faster. Runtime menu at the top right, change runtime type, select T4.

Step 4: Install Dependencies

Run the dependency installation cell. This sets up all the libraries you need. It takes a few minutes the first time.

Step 5: Configure for Remote Access

Run the Ngrok configuration cell. This deploys a web UI you can access from your browser. Without this step, the interface only runs inside the Colab environment where you can't interact with it.

Step 6: Get Your API Keys

You need two free API keys:

  • Gemini API key (free tier available) — grab yours at this walkthrough if you need help
  • Pexels API key (free tier available) — sign up at pexels.com/api

Paste both into their respective cells and run them.

Step 7: Access the Web UI

Once configured, you'll get a URL in the output. Click it and press "Visit Site" to open the full web interface.

Step 8: Configure Basic Settings

In the web UI, open basic settings. Select Gemini as your LLM provider and paste your API key. Set Pexels as your footage source and paste that API key too.

Step 9: Enter Your Video Topic

Input the topic for your video. Example: "How ordinary people are making their first $10K online without selling anything"

Be specific with your topic. Vague topics produce vague scripts.

Step 10: Generate Script and Keywords

Click "Use AI to generate video script and video keywords." Gemini will automatically write both the script and the search terms it'll use to pull matching footage from Pexels.

Step 11: Choose Your Settings

  • Select portrait (9:16) for TikTok/Reels or landscape (16:9) for YouTube
  • Pick your voiceover (example: American English)
  • Leave background music on random or upload custom audio

Step 12: Generate the Video

Click generate and wait. It takes around 20 minutes. Go grab coffee. The tool is assembling the script, pulling footage, generating the voiceover, adding subtitles, and mixing in background music — all automatically.

Step 13: Watch and Download

Once done, watch your finished video in the UI. Download it and upload directly to your platform.

Final Thoughts

No script writing. No footage hunting. No editing. One topic, 20 minutes later, you have a publishable video.

Everything here is free. Limited usage on the free tiers, but more than enough to test and iterate before committing to a paid plan. Watch the full walkthrough below to see it in action.