Maestro
Maestro is a local, one-click AI video, image and music studio built on the WanGP pipeline, with a director mode that plans a full music video or short film, screenplay, shots and clips, from a single prompt.
Maestro combines image, video and audio generation models with an LLM-directed production workflow and a non-destructive multi-track editor, aimed at LTX-2.5, MiniMax H3 and other current local generative models and LoRAs. It bundles Qwen Image 2.1 7B for unified image generation and editing with reference images and transparent PNG output.
Its director mode takes an audio track or a written story and has a local LLM plan every shot, write the screenplay or lyrics, generate start frames and character-consistent keyframes, adapt prompts to each model's requirements, and run the full multi-clip generation. A music video skill analyzes tempo, song sections and energy to align shots to the beat and can separate and label different singers; a short film skill writes scenes with named characters, dialogue and continuity across cuts. An automatic mode runs the whole pipeline end to end, while a manual mode allows review at each step.
A performance auto-tune step detects the available GPU, VRAM and RAM on first launch and selects a matching quantization and memory profile, with manual override still available. It installs through Pinokio.