Automating the short-video content assembly line

What it does

MoneyPrinterTurbo is a Python application that treats short-video creation like a factory assembly line. Feed it a topic or keyword, and it prompts an LLM to write a script, pulls copyright-free HD video clips (or uses your own local files), generates a voiceover via TTS, burns in styled subtitles, layers background music, and exports a finished cut in either 9:16 vertical or 16:9 horizontal resolution. It exposes both a web interface and a REST API, so you can click through a UI or call it programmatically.

The interesting bit

The project integrates with more than a dozen LLM providers—from OpenAI and DeepSeek to local Ollama instances and free tiers like gpt4free—so you are not locked into a single API. It also defaults to free Edge TTS for narration and offers a choice between fast, lightweight subtitle timing or slower, more accurate local faster-whisper transcription, letting you trade precision for speed depending on your hardware.

Key highlights

• Supports 13+ LLM backends including local models via Ollama and cloud APIs like Azure, Gemini, and DeepSeek
• Outputs both 9:16 (1080×1920) vertical and 16:9 (1920×1080) horizontal HD video
• Batch generation mode produces multiple candidate videos from one prompt
• Free Edge TTS voice synthesis by default, with optional Azure TTS V2 for higher-quality narration
• Two subtitle engines: fast Edge TTS timestamps or local faster-whisper for finer-grained alignment

Caveats

• The fast subtitle mode can misalign timestamps on complex sentences, while the slower Whisper mode drags on CPU-only machines
• The Windows one-click installer still ships v1.2.6 and needs a manual patch to reach current code
• The README explicitly warns against Chinese file paths and global-tunnel VPNs, hinting at environmental fragility

Verdict

Developers building automated content pipelines or creators who need to prototype short-form video at volume should look here; anyone seeking handcrafted cinematic editing or who balks at AI voiceover should keep walking.