Pallaidium
PALLAIDIUM — a generative AI movie studio, seamlessly integrated into the Blender Video Editor (VSE), enabling end-to-end production from script to screen and b
What is Pallaidium?
PALLAIDIUM — a generative AI movie studio, seamlessly integrated into the Blender Video Editor (VSE), enabling end-to-end production from script to screen and back.
Overview of Pallaidium
What it is
Pallaidium is a free generative AI movie studio integrated into Blender's Video Sequence Editor (VSE). It allows end-to-end production from script to screen and back, using AI to prototype the full production cycle before committing to real footage. It is open source under GPL-3.0.
Key features
- Text to video, text to image, text to text, text to speech, text to audio, text to music - Image to image, image to video, image to text - Video to video, video to image, video to text - ControlNet, OpenPose, Canny, ADetailer, IP Adapter Face, IP Adapter Style - Multiple LoRAs, LoRA Weight, Style selector, Seed, Quality steps, Strip power, Frames (Duration), Word power, Model card selector - Batch conversion, batch refinement of images, prompt batching, batch upscale & refinement of movies - Render-to-path selector, render finished notification, Render Queue (async batch jobs), job status tracking (PENDING / RUNNING / COMPLETED), Start / Stop / Cancel controls - User-defined file path for generated files, seed and prompt added to strip name, one-click install and uninstall dependencies
Use cases
- Prototyping a film: start from song lyrics or scene description, convert to screenplay with shots, break down into timed strips, generate images, animate, add voices and music, edit based on emotional weight.
- Reverse process: extract captions from result, reconstruct screenplay, refine, then shoot real film with confidence.
- Generating media directly from text prompts or from existing strips in the VSE timeline.
Getting started
Requirements: Windows (limited Linux support), Blender 5.2 or later, CUDA-supported Nvidia card with 6-16 GB VRAM, CUDA 12.8, 20+ GB HDD. Installation: 1. Download and install git (must be on PATH): https://git-scm.com/downloads 2. Download Blender 5.2 or later from https://www.blender.org/download/ and unzip into Documents folder (avoid spaces/special characters in path). 3. Download the add-on from https://github.com/tin2tin/Pallaidium/archive/refs/heads/main.zip 4. Install add-on: Preferences > Add-ons > Install > select file > enable. 5. In Generative AI add-on preferences, click 'Install Dependencies'. 6. Restart computer, then open Blender. 7. Open add-on UI in Sequencer > Sidebar > Generative AI. 8. First model execution downloads 5-30 GB.
License and pricing
License: GPL-3.0. Pricing: open source (free).
Based on the project's public README and repository metadata.
