Blizaine/Maestro
An all-in-one, 100% local AI video, image, and music studio. Director mode plans full music videos and short films from a single prompt. Built on the WanGP pipeline. Install via Pinokio.
What it solves
Maestro is a local AI creative studio designed to simplify the complex process of producing AI-generated videos, music videos, and short films. It removes the technical friction of managing multiple AI models, prompt engineering, and hardware optimization, providing a unified interface for creators to go from a story idea or audio track to a finished, edited video.
How it works
Maestro integrates various local generative AI models for image, video, and audio. It features a "Director Mode" where a local LLM (like Gemma 4) acts as a production lead, planning shots, writing screenplays, and managing character consistency across clips. The system uses a multi-pass refinement process to turn high-level ideas into model-specific prompts. For manual control, "Studio Mode" provides direct access to model parameters, LoRAs from CivitAI, and advanced tools like face refining and video blending. Finally, an "Editor Mode" allows users to arrange these generated assets on a non-destructive multi-track timeline for final polishing.
Who it’s for
It is built for digital creators, filmmakers, and AI artists who want to produce high-quality, long-form AI video content locally without needing to manage separate tools or complex cloud services.
Highlights
- Director Mode: Automatically generates music videos (beat-aware) and short films (screenplay-driven) using an LLM-directed pipeline.
- Performance Auto-Tune: Automatically detects hardware specs to optimize VRAM and RAM settings for a zero-config setup.
- Integrated Editor: A built-in multi-track timeline for trimming, splitting, and arranging AI-generated clips.
- CivitAI Integration: A built-in browser to search, install, and apply LoRAs with AI-generated prompting guides.
- Character Consistency: Tools for saving and sharing portable characters with embedded appearance and voice data.
- Local LLM: Built-in support for GGUF models (e.g., Gemma 4) to handle planning and prompt enhancement without external APIs.
Related
- Project
- Project
- Project
- Project
- Project