Maestro: video generator — what it does and what you need to run it

Maestro is an open-source video generator by Blizaine, positioned as a local studio for creating videos, images, and music. It works on the image-text-to-video scheme: text and optionally an image are fed in, and a video sequence comes out. Installation is done through the Pinokio platform.

What it does

As described by the authors, Maestro combines video, image, and music generation in one tool. Director mode allows planning full music videos and short films from a single text prompt. The video generation is based on the WanGP pipeline. For the language model, the minimax-h3 library is used, and the model is loaded from huggingface.co/MiniMaxAI/MiniMax-H3.

The authors note that Maestro automatically downloads llama-server (about 600 MB, once) and the selected GGUF model on first launch. By default, Gemma 4 4B is offered — according to the developer, it runs fast and comfortably even on small GPUs. The system itself detects CUDA and binds the language model to the GPU if available.

What you need to run it

The platform, as described by the author, is CUDA. VRAM requirements: 12–24 GB; this is the developer's claim, not the result of independent measurements. The description mentions that after an update, cards with 12 GB or more get a fast language model engine; previously, an audio profile mistakenly blocked it on cards with less than 24 GB of memory.

The Python version, as described by the author, is 3.11. The developer states that the tested environment of Python 3.11 / PyTorch 2.10 / CUDA 13 has become standard for installation, updates, and running on compatible systems with RTX 40-series and 50-series GPUs, while fallback options for other configurations are retained.

The largest model weights file takes 9.7 GB, and the total size of all model weights files is 464.1 GB. The code license is non-standard, and the authors recommend checking it manually. The model weights license is marked as other. The code is written in Python, and the repository is available at github.com/Blizaine/Maestro. The latest code change was on September 4, 2026, and the repository was created on July 8, 2026.

Who it suits

The tool is designed for those who need fully local video generation without relying on cloud services. Director mode may be useful for creators assembling music videos or short films from a single script. The open source allows understanding the pipeline structure and adapting it to your own tasks if needed.

It won't suit those without a CUDA-supported GPU and sufficient VRAM: as described by the author, 12 GB or more is needed for comfortable work. The total model weights size of 464.1 GB requires significant disk space if you download all files. The non-standard code license and the other model weights license mean you should carefully review the terms before commercial use.

Maestro is a local open-source video generator that combines multiple types of generation in one application. It targets users with CUDA-compatible GPUs and sufficient memory, and installation via Pinokio simplifies deployment. Before use, you should separately check the terms of the code and model weights licenses.