Kimodo Motion Diffusion
Generate text-driven 3D human or humanoid robot motions for sim demos, keyframe-constrained animation, and export to AMASS or qpos without a full mocap pipeline.
Overview
Kimodo Motion Diffusion is an agent skill for the Validate phase that generates 3D human and humanoid robot motions with Kimodo using text prompts and kinematic constraints.
Install
npx skills add https://github.com/aradotso/trending-skills --skill kimodo-motion-diffusionWhat is this skill?
- Text-prompt and kinematic-constraint control (full-body keyframes, end-effector pose, 2D paths and waypoints)
- Human and humanoid robot motions from a diffusion model trained on 700 hours of commercial-friendly optical mocap
- CLI entry points kimodo_gen and kimodo_demo plus optional Docker workflow with Gradio on port 7860
- Export paths to AMASS or qpos for downstream sim and tooling
- Automatic Hugging Face model download on first run (~17GB VRAM GPU recommended)
- 700 hours of commercially-friendly optical mocap in the training set
- ~17GB VRAM recommended for generation
- Models download automatically on first use from Hugging Face
Adoption & trust: 1.2k installs on skills.sh; 31 GitHub stars; 2/3 security scanners passed (skills.sh audits).
What problem does it solve?
You need realistic body motion for a sim or demo but lack mocap, animation staff, or time to hand-author keyframes.
Who is it for?
Indie builders prototyping humanoid robots, game characters, or embodied AI demos who already have a Linux GPU box or Docker with ~17GB VRAM.
Skip if: Teams shipping a standard web SaaS with no 3D/sim pipeline, or anyone who cannot run GPU workloads locally or in Docker.
When should I use this skill?
Generate human motion with text prompt, create robot animation with diffusion model, add keyframe or end-effector constraints, generate G1 robot motion for MuJoCo, or export to AMASS/qpos.
What do I get? / Deliverables
You get generated 3D motions you can preview in the demo CLI and export in AMASS or qpos formats for MuJoCo-style validation.
- Generated 3D motion sequences from prompts and constraints
- Exports in AMASS or qpos format
- Optional interactive demo session via kimodo_demo
Recommended Skills
Journey fit
Motion diffusion is most often pulled in when you need a believable movement prototype to validate a robotics, avatar, or game concept before investing in custom animation or hardware. Prototype subphase fits early demos (walking cycles, G1/MuJoCo clips, constrained end-effector paths) that prove feasibility rather than shipping production runtime integration.
How it compares
Use for diffusion-based motion synthesis with constraints, not for generic LLM text generation or a hosted video API integration.
Common Questions / FAQ
Who is kimodo-motion-diffusion for?
Solo builders and small teams validating robotics, animation, or game concepts who want agent-guided setup and prompts for Kimodo motion generation.
When should I use kimodo-motion-diffusion?
Use it during Validate when you need walking or manipulation clips, MuJoCo G1 motions, keyframe- or end-effector-constrained paths, or an interactive kimodo_demo before committing to a custom animation pipeline.
Is kimodo-motion-diffusion safe to install?
Review the Security Audits panel on this Prism page and treat the upstream Kimodo repo, pip install, Docker image, and Hugging Face model downloads as third-party supply-chain risk before running on sensitive machines.
SKILL.md
READMESKILL.md - Kimodo Motion Diffusion
# Kimodo Motion Diffusion > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. Kimodo is a kinematic motion diffusion model trained on 700 hours of commercially-friendly optical mocap data. It generates high-quality 3D human and humanoid robot motions controlled through text prompts and kinematic constraints (full-body keyframes, end-effector positions/rotations, 2D paths, 2D waypoints). ## Installation ```bash # Clone the repository git clone https://github.com/nv-tlabs/kimodo.git cd kimodo # Install with pip (creates kimodo_gen and kimodo_demo CLI commands) pip install -e . # Or with Docker (recommended for Windows or clean environments) docker build -t kimodo . docker run --gpus all -p 7860:7860 kimodo ``` **Requirements:** - ~17GB VRAM (GPU: RTX 3090/4090, A100 recommended) - Linux (Windows supported via Docker) - Models download automatically on first use from Hugging Face ## Available Models | Model | Skeleton | Dataset | Use Case | |-------|----------|---------|----------| | `Kimodo-SOMA-RP-v1` | SOMA (human) | Bones Rigplay 1 (700h) | General human motion | | `Kimodo-G1-RP-v1` | Unitree G1 (robot) | Bones Rigplay 1 (700h) | Humanoid robot motion | | `Kimodo-SOMA-SEED-v1` | SOMA | BONES-SEED (288h) | Benchmarking | | `Kimodo-G1-SEED-v1` | Unitree G1 | BONES-SEED (288h) | Benchmarking | | `Kimodo-SMPLX-RP-v1` | SMPL-X | Bones Rigplay 1 (700h) | Retargeting/AMASS export | ## CLI: `kimodo_gen` ### Basic Text-to-Motion ```bash # Generate a single motion with a text prompt (uses SOMA model by default) kimodo_gen "a person walks forward at a moderate pace" # Specify duration and number of samples kimodo_gen "a person jogs in a circle" --duration 5.0 --num_samples 3 # Use the G1 robot model kimodo_gen "a robot walks forward" --model Kimodo-G1-RP-v1 --duration 4.0 # Use SMPL-X model (for AMASS-compatible export) kimodo_gen "a person waves their right hand" --model Kimodo-SMPLX-RP-v1 # Set a seed for reproducibility kimodo_gen "a person sits down slowly" --seed 42 # Control diffusion steps (more = slower but higher quality) kimodo_gen "a person does a jumping jack" --diffusion_steps 50 ``` ### Output Formats ```bash # Default: saves NPZ file compatible with web demo kimodo_gen "a person walks" --output ./outputs/walk.npz # G1 robot: save MuJoCo qpos CSV kimodo_gen "robot walks forward" --model Kimodo-G1-RP-v1 --output ./outputs/walk.csv # SMPL-X: saves AMASS-compatible NPZ (stem_amass.npz) kimodo_gen "a person waves" --model Kimodo-SMPLX-RP-v1 --output ./outputs/wave.npz # Also writes: ./outputs/wave_amass.npz # Disable post-processing (foot skate correction, constraint cleanup) kimodo_gen "a person walks" --no-postprocess ``` ### Multi-Prompt Sequences ```bash # Sequence of text prompts for transitions kimodo_gen "a person stands still" "a person walks forward" "a person stops and turns" # With timing control per segment kimodo_gen "a person jogs" "a person slows to a walk" "a person stops" \ --duration 8.0 --num_samples 2 ``` ### Constraint-Based Generation ```bash # Load constraints saved from the interactive demo kimodo_gen "a person walks to a table and picks something up" \ --constraints ./my_constraints.json # Combine text and constraints kimodo_gen "a person performs a complex motion" \ --constraints ./keyframe_constraints.json \ --model Kimodo-SOMA-RP-v1 \ --num_samples 5 ``` ## Interactive Demo ```bash # Launch