Vídeo
Wan2GP
A lean, fast video generator tuned to squeeze big results out of modest GPUs.
★ 9,9 mil estrelasCustom (WanGP Community 2.0)
Ver o guia →
// dados do projeto
Dados do projeto
- Repositório
- Blizaine/Maestro
- Estrelas no GitHub
- 686
- Forks
- 120
- Licença
- Custom (non-commercial)
- Última atualização no GitHub
- 28 de setembro de 2026
- Categoria
- Vídeo
- Tipo de ficha
- Guia de instalação completo
- Site oficial
- Install via Pinokio
// como funciona
Como funciona
- Install it in one click
Get Pinokio, open its Discover tab, search for “Maestro” and hit Install. The launcher builds a hardware-matched Python environment and the app UI automatically — allow 10 to 20 minutes.
- Choose Director or Studio
Director mode takes a song or story and plans the whole production for you. Studio mode gives you direct control over models, prompts, and parameters.
- Generate keyframes, then clips
The pipeline creates still keyframes first so you can approve the look, then renders each clip — with beat-aware timing for music videos.
- Polish and export
Re-roll any single clip without redoing the whole run, apply prompt polish passes, combine everything, and export your finished piece.
// recursos
Por que as pessoas gostam tanto
Director Mode
An LLM writes the screenplay, plans shots, and times cuts to your music's BPM, sections, and energy — with named, consistent characters.
Studio Mode
Manual control over video, image, and audio models, with multi-clip pipelines and shared LoRAs for long-form work.
Hardware auto-tune
Detects your GPU, VRAM, and RAM on first launch and picks the right performance profile — no settings guesswork.
CivitAI LoRA browser
Search, filter, and one-click install LoRAs inside the app, with update badges and AI-written prompt guides per LoRA.
OOM recovery
If a generation runs out of VRAM, a banner offers a one-click fix so you can lower headroom and resume instead of crashing.
AI edit modes
Retake a section with a new prompt, outpaint frames in any direction, or modify elements in existing footage.
Advertisement
// execute você mesmo
Execute você mesmo
- Step 1 — Install Pinokio
Download and install Pinokio from pinokio.computer — it is the one-click launcher Maestro ships through.
- Step 2 — Install Maestro
In Pinokio, open the Discover tab, search for “Maestro”, and click Install.
- Step 3 — Start
When the install finishes, click Start. Your first generation in each model triggers a one-time weight download, then everything runs locally.
// casos de uso
O que você pode criar
- Produce a full music video timed to your track's downbeats.
- Draft a short film with consistent characters and dialogue.
- Experiment with LoRAs and video models without manual environment setup.
- Re-roll and polish individual clips inside a bigger production.
Advertisement
// perguntas frequentes
Perguntas frequentes
Is Maestro really free?
The app itself is free to download and use, and there is no subscription. Do note it ships under a custom non-commercial evaluation license (based on the WanGP license), so read the LICENSE file in the repo before any commercial use.
What hardware do I need?
A Windows or Linux PC with an NVIDIA GPU is the supported setup. On first launch Maestro detects your GPU, VRAM, and RAM and tunes itself accordingly. macOS support is in development.
Does it need an internet connection?
Only for setup and the first model-weight download. All generation happens locally on your machine — your prompts and files never leave it.
How is Maestro different from Wan2GP?
Maestro is built on top of the WanGP/Wan2GP pipeline and adds a full studio interface, Director mode, LoRA management, and editing tools around it.
Can I use my own models and LoRAs?
Yes — the built-in CivitAI browser handles discovery and one-click installs, and you can point it at your own checkpoints and LoRAs too.
// continue explorando
Ferramentas semelhantes
Vídeo
Open Generative AI
A self-hosted creative suite with 600+ image and video models behind one clean interface.
★ 29,6 mil estrelasMIT
Ver o guia →Imagem
ComfyUI
The node-graph powerhouse for precise, professional AI image and video generation.
★ 136,0 mil estrelasGPL-3.0
Ver o guia →Vídeo
locally-uncensored
The all-in-one local AI studio for your desktop: chat, image and video generation and a coding agent in one free, open source app. Windows and Linux. No Dock…
★ 1,9 mil estrelasAGPL-3.0
Ver o guia →Vídeo
VideoLingo
Netflix-level video subtitles and dubbing - a one-click AI subtitle team
★ 18,7 mil estrelasApache-2.0
Ver o guia →Vídeo
ai-moive-studio
Infinite-canvas AI video studio — script to distribution in one workflow.
★ 1,6 mil estrelasApache-2.0
Ver o guia →// alternativas
Alternativas em Vídeo
Os outros projetos de Vídeo com mais estrelas no catálogo do LibreAI — a lista realista de alternativas se Maestro não for o ideal para você.
- yt-dlp★ 195,3 mil
A feature-rich command-line audio/video downloader
- Deep-Live-Cam★ 96,9 mil
real time face swap and one-click video deepfake with only a single image
- obs-studio★ 76,9 mil
OBS Studio - Free and open source software for live streaming and screen recording
- FFmpeg★ 64,7 mil
Mirror of https://git.ffmpeg.org/ffmpeg.git
- OpenMontage★ 62,7 mil
World's first open-source, agentic video production system. 12 production pipelines, 100+ tools, 700+ agent skill and production-knowledge files. Turn your A…
// mais em video
Mais em Vídeo
Projetos de Vídeo próximos no catálogo, os mais próximos em estrelas de Maestro: continue explorando dos dois lados dele.
- podcast-maker★ 707
Turn newsletters into daily YouTube videos — motion graphics plus TTS.
- NotchPrompter★ 645
A floating always-on-top teleprompter for macOS — perfect for quick videos.
- kandinsky-5★ 829
Open diffusion models for both video and image generation.
- BiliSum★ 620
AI video summarizer and knowledge base for Bilibili, YouTube and local videos.
- ConsisID★ 862
Text-to-video that keeps identities consistent across every frame.
- t2v_metrics★ 603
Evaluate text-to-video models with VQAScore-based metrics.
Baixe o Maestro
Grátis e aberto: pegue o código-fonte, dê uma estrela no repositório e comece a criar.