Before I built my own pipeline I looked hard at the tools everyone links to. They're real. The promise around them isn't.
What they actually are
- MoneyPrinterTurbo (~99.8k stars, MIT) — scrape a topic, generate a script, add TTS and stock footage, output a video.
- OpenMontage (~43.4k stars, AGPL-3.0) — a similar montage/automation stack.
Both are genuine, working projects. If your goal is "turn a prompt into a finished clip with no editing," they deliver. I evaluated them seriously.
Why I didn't use either
Two concrete reasons, neither about code quality:
- The AGPL trap. OpenMontage is AGPL-3.0. Wrap it in a web service and the copyleft can reach your whole service. For anything I might expose, that's a risk I won't take on casually.
- Wrong assets for my job. My highest-value use is promoting my own apps, where the video should show my app's real screens — not random stock footage. MoneyPrinter's stock-and-LLM pipeline is optimized for the opposite. So I built a small pipeline around my own brand assets instead.
The honest part: production was never the bottleneck
The word "money printer" implies the hard part is making the videos. It isn't. I can already render a finished Short in seconds. What stands between output and results is reach — the algorithm, the niche, and platform authenticity rules that specifically target template-farmed, low-transformation output.
I only know this because I published a small batch and read the numbers: reach was fine for some niches and near-zero for others, and retention only held where the video delivered its payoff on screen. A tool that outputs 50 clips does not output 50 clips' worth of views.
Use these projects if they fit — they're legitimate. Just don't buy the framing. Automating production lowers the cost of trying; it doesn't print money. The money, if it comes, is on the distribution side the repos can't automate for you.