What is MoneyPrinterTurbo?
MoneyPrinterTurbo takes a topic or keyword and produces a finished short video: a language model writes the script, it matches footage, adds a voiceover, subtitles and background music, and renders portrait 9:16 (1080x1920), landscape 16:9 or square 1:1 output. You can also supply your own script, and finished videos can be published to TikTok, Instagram and YouTube Shorts.
It works with a long list of providers: OpenAI, Anthropic, Gemini, DeepSeek, Qwen, Kimi and others directly, OpenRouter and LiteLLM gateways, and Ollama for local models. Footage comes from Pexels, Pixabay and Coverr stock, your own files, or AI video services such as MiniMax H3 and Seedance. Voice options start with free Edge TTS and extend to Azure, ElevenLabs, Kokoro, Fish Audio and more; subtitles use TTS timestamps or local faster-whisper.
It runs as a web UI, a REST API, a CLI with batch files, or an agent skill. It is written in Python 3.11+, MIT licensed, and installs with pip install mpt, Docker, or a one-click Windows package. The project started in March 2024.
Is MoneyPrinterTurbo worth it?
MoneyPrinterTurbo has become the default open-source answer to automated short video: 139,673 stars, 22,273 forks and steady releases since March 2024, the latest v1.3.9 on 21 October 2026. The reason is breadth. Almost every LLM provider, gateway and local runtime is supported, as are a dozen voice engines and several AI video services, so you can assemble a pipeline from whatever you already pay for.
The cheapest configuration costs very little. Pexels and Pixabay footage are free, Edge TTS is free with no API key, and the default subtitles use TTS timestamps with no GPU. The README's minimum is 4 CPU cores and 4 GB of RAM with no GPU; a GPU of 4 GB or more only helps if you switch subtitles to local faster-whisper, whose default large-v3 model is about 3 GB. Swap stock footage for AI clips and the cost moves to the video provider: at Metaso's 768p rate of CNY 0.09 a second, 30 seconds of generated footage is CNY 2.70.
The catch is the result. Stock clips matched by keyword to a machine-written script produce the generic faceless video that platforms are full of, and the README itself says the bundled background music came from YouTube and should be deleted if copyright is a concern. For a brand channel the pipeline is a draft generator; the newer option to replace individual scenes and rebuild only affected stages helps, but someone still has to edit.
Operationally it is straightforward. The web UI binds to 127.0.0.1:8501 and the API to port 8080 by default; setting MPT_WEBUI_HOST to 0.0.0.0 opens it to the network, and the README describes no login, so keep it behind your own access control. Batch runs accept up to 100 tasks per file and carry on past individual failures. Note that the PyPI package named MoneyPrinterTurbo is run by a third party; the official one is mpt.
Where it earns its place
A one-person faceless explainer channel
Turn a list of topics into draft Shorts every day for almost nothing
Write the topics to a JSONL file and run cli.py --batch-file with DeepSeek or a local Ollama model, Pexels footage and Edge TTS; up to 100 videos per batch, with a summary of successes and failures. Spend is LLM tokens only.
A developer building a marketing tool for restaurants and gyms
Add generate-a-promo-video to our product without building a video pipeline
Run the Docker image and call the REST API documented at port 8080/docs with a script and settings; keep the API on a private network, since no login is described.
A small e-commerce brand's content lead
Produce product videos with on-brand footage instead of stock
Upload your own product clips as local material, write the script yourself, pick an ElevenLabs or Azure voice, and use scene replacement to fix individual shots. Paid TTS and any AI clips are billed by those providers.
Before you commit
- Setup
- pip install mpt then mpt webui on Python 3.11, docker compose with the ghcr.io image after copying config.example.toml, or the Windows .7z package (update.bat, then start.bat).
- Integrations
- OpenAI, Anthropic, Gemini, DeepSeek, Qwen, Kimi, Azure OpenAI, OpenRouter, LiteLLM, Ollama and more; Pexels, Pixabay, Coverr, MiniMax H3, Seedance; Edge TTS, Azure, ElevenLabs, Kokoro, Fish Audio; faster-whisper; publishing to TikTok, Instagram and YouTube Shorts.
- Data and privacy
- Self-hosted; scripts go to your chosen LLM provider and text to your TTS provider. Web UI listens on 127.0.0.1 unless MPT_WEBUI_HOST is changed. MIT licence; bundled music may carry copyright.
- Cost at scale
- Runs on a 4-core, 4 GB machine with free stock and Edge TTS; costs scale with LLM tokens, paid voices and AI footage, which at CNY 0.15 a second for 2K MiniMax H3 dominates everything else.
- Limits
- Output quality depends on keyword-to-footage matching; batch files capped at 100 tasks and 1 MiB; ffmpeg path and open-file limits are common setup problems; 71 open issues and pull requests.
What’s good
- Supports nearly every LLM provider plus Ollama for local models
- Runs free on stock footage and Edge TTS with no GPU
- Web UI, REST API, CLI batch mode and agent skill
- Docker image, pip package and one-click Windows build
- Active since March 2024 with regular releases
What’s not
- Keyword-matched stock footage looks generic
- Bundled music has unclear copyright
- No login described for the web UI or API
- Third-party PyPI package with a confusingly similar name
Best for
Developers and small media teams who want a self-hosted, scriptable short-video generator built from providers they already use, and will edit the output.
MoneyPrinterTurbo alternatives
MoneyPrinterTurbo: common questions
- Is MoneyPrinterTurbo worth it?
- MoneyPrinterTurbo is a mature, provider-agnostic way to churn out faceless short videos at almost no cost, and the output looks exactly as templated as that sounds. We rate it 4.0 out of 5.
- How much does MoneyPrinterTurbo cost?
- Free under MIT. Costs are the services you plug in: LLM tokens, optional paid TTS, and AI video clips such as Metaso MiniMax H3 at CNY 0.09 per second for 768p or CNY 0.15 for 2K. Stock footage and Edge TTS are free.
- What are the downsides of MoneyPrinterTurbo?
- Keyword-matched stock footage looks generic; bundled music has unclear copyright; no login described for the web UI or API; third-party PyPI package with a confusingly similar name.
- Who is MoneyPrinterTurbo best for?
- Developers and small media teams who want a self-hosted, scriptable short-video generator built from providers they already use, and will edit the output.
- How much does MoneyPrinterTurbo cost for a team or in production?
- Runs on a 4-core, 4 GB machine with free stock and Edge TTS; costs scale with LLM tokens, paid voices and AI footage, which at CNY 0.15 a second for 2K MiniMax H3 dominates everything else.
- Is MoneyPrinterTurbo safe for company data?
- Self-hosted; scripts go to your chosen LLM provider and text to your TTS provider. Web UI listens on 127.0.0.1 unless MPT_WEBUI_HOST is changed. MIT licence; bundled music may carry copyright.
- What are the best MoneyPrinterTurbo alternatives?
- The closest alternatives we have reviewed are VideoAgent.
Answers checked Nov 5, 2026.
Make MoneyPrinterTurbo? Show this rating on your site
