Video2rig

Video2rig: Video tool that turns AI-generated model videos into Blender skeleton animations without hand-placed points.

Video2rig turns an AI video of your own 3D model into an animation for your own rig inside Blender. You render a start picture of your rigged model, use an image-to-video generator to animate it from that first frame, and video2rig rebuilds that motion on your skeleton. The tool trains a small joint reader on random pose renders of your model and then reads every clip joint by joint. It supports custom rigs, Mixamo, and auto-generated skeletons, and it tracks props like logs and stones alongside characters. The software runs locally on your computer with NVIDIA graphics cards and requires Blender, Python, and PyTorch. It is made for developers and animators who want to convert AI video clips into editable Blender Actions.

Video2rig at a glance

Best for: Software engineer, Designer
Runs on: Windows, Linux, self-hosted, open source
Pricing: Free
Released: October 2026
Category: AI video tool
Website: github.com/canitcode/video2rig

Key features

Rebuilds motion frame by frame onto custom or standard skeletons
Trains a small joint reader on random model poses
Auto-rigs models without existing skeletons using UniRig
Tracks rigid props alongside characters
Supports one or two cameras in a single frame
Outputs ordinary Blender Actions for editing and export
Runs locally on your own computer without cloud dependence

Use it for

Converting AI character video animations into editable Blender actions
Animating custom 3D creature models with AI video generators
Rigging unrigged 3D meshes automatically for game development
Tracking character interactions with held props like weapons and logs

Alternatives to Video2rig

Prompt2rig — Generates jitter-free 3D character motion from text prompts via video models and local GPU passes.
img2threejs — Generates code-only, animation-ready Three.js models from a single image.
Img2interior — Generates editable Blender scenes and interactive Three.js viewers from one room image.
toyxyz RGBMocap — Runs AI models locally on Windows to turn video into editable 3D character animation.
Toonflow — Automates the entire video creation workflow with multi-agent collaboration and character consistency.
Motion Skills — Generates professional motion graphics from text without manual keyframing.
video2ctx — Provides video context with timestamped citations and uses 95% fewer tokens than transcripts.
LightX2V — It combines avatars, image-to-video, and image generation in one workspace.
Motionmaxxing — It replaces default AI templates with measured motion based on professional design rules.
Pixel2Motion — Generates clean semantic SVGs and deterministic motion QA evidence from static images.
ManimGX — Renders Manim CE compatible scenes up to 90 times faster with native and WebGPU support.
Kinetograph — Renders every frame on the GPU without a browser.

All Video2rig alternatives