Multi-input asset creation
Upload raw 3D models, 2D images, or describe the asset in text. The site says no pre-processing or manual format conversion is required before upload.
Rigyd
Rigyd generates physically accurate SimReady 3D assets and worlds for robotics simulation from 3D models, images, or text. Exports OpenUSD and MJCF for Isaac Sim and MuJoCo.
Overview
Rigyd generates SimReady 3D assets and worlds for robotics simulation. It takes raw 3D models, images, or text descriptions and produces physically accurate output that can be used in simulation workflows instead of geometry-only files.
The published pages emphasize automatic enrichment for simulation: physics properties, collision geometry, semantic labeling, and simulator-friendly packaging in OpenUSD and MJCF. Rigyd positions this workflow for teams working in NVIDIA Isaac Sim, MuJoCo, Genesis, and related robotics environments.
Core capabilities
Physics property estimation
Rigyd estimates density, friction, restitution, mass, center of mass, and inertia from the source content, using mesh volume, geometry, and a calibrated materials database.
Collision mesh generation
The pipeline generates collision geometry automatically with convex decomposition/V-HACD, which the site presents as optimized for simulation runtimes like PhysX.
Simulation-ready exports
Export SimReady assets in OpenUSD with USDPhysics schemas and in MJCF for MuJoCo. The homepage also says assets are Z-up, metric, and single rigid bodies, with URDF coming soon.
Validation and inspection
A 106-check validator and Omniverse asset validation report are part of the output flow, helping confirm structure before the asset is used in simulation.
Batchable world generation
The product is built to produce large volumes of distinct assets and scenes, including domain-randomization-friendly variations for robotics training pipelines.
Where Rigyd fits
Turn raw assets into SimReady files
Convert existing CAD or mesh libraries into physically accurate simulation assets when you need more than just rendered geometry.
Build training data at scale
Generate large collections of distinct objects and scene variations for domain randomization and robot policy training.
Feed Isaac Sim workflows
Prepare Isaac Sim-compatible OpenUSD with physics schemas, collision geometry, and validation for robotics teams using NVIDIA’s stack.
Produce MuJoCo assets
Create MuJoCo-ready MJCF assets when you need physics-aware objects for robot learning, testing, or evaluation.
Automate high-volume asset pipelines
Support enterprise simulation programs that need API-driven batch creation of many assets and scenes across a catalog or warehouse-like environment.
Pros and Cons
Pros
- Accepts raw 3D files, images, and text, reducing the need for manual preprocessing.
- Automatically estimates physics properties such as mass, friction, restitution, center of mass, and inertia.
- Generates collision meshes and exports validated OpenUSD and MJCF assets ready for simulation use.
- Includes validation steps and an Omniverse validation report, which can help catch structural issues before deployment.
- Supports batch processing and API access for larger asset volumes, which suits enterprise simulation pipelines.
Cons
- The pricing page is not available, so the public site does not show standard plans or pricing details.
- URDF is listed as coming soon, so teams that rely on URDF may need a different output path today.
- The strongest evidence is for Isaac Sim and MuJoCo; support for other simulators is mentioned more lightly on the site.
FAQ
What inputs does Rigyd accept?
Rigyd accepts 3D models such as .glb, .fbx, and .obj, plus 2D images and text descriptions. The source pages say no pre-processing is needed before upload; the system handles geometry cleanup, physics estimation, collision mesh generation, and semantic labeling automatically.
What formats does Rigyd export?
Rigyd outputs OpenUSD assets with USDPhysics schemas and MJCF exports. The Isaac Sim page says the OpenUSD output includes PhysicsRigidBodyAPI, PhysicsMassAPI, and PhysicsCollisionAPI, and the homepage says URDF is coming soon.
How does the workflow work?
The source pages describe a fully automated pipeline. You upload a model, image, or text prompt; Rigyd renders views, estimates physics properties from geometry and materials, generates collision hulls, and then validates the result before download.
Who is Rigyd for?
Rigyd is positioned for robotics simulation, embodied AI, and robot learning/evaluation workflows that need physically accurate assets and worlds. The site specifically mentions Isaac Sim, MuJoCo, Genesis, and Gazebo as supported simulator targets or related pages.
How is Rigyd priced?
The pricing page on the site is unavailable, but the homepage and Isaac Sim page indicate API access and an Enterprise plan for batch processing. The site also says to request API access for higher-volume use.
Quick Facts
- Category
- Robotics simulation / 3D asset generation
- Website
- rigyd.com
- Primary outputs
- OpenUSD and MJCF
- Typical inputs
- 3D models, images, text descriptions
- Primary simulators
- NVIDIA Isaac Sim, MuJoCo, Genesis
- Access model
- API access / enterprise-oriented batch processing
Rigyd Alternatives
DeepMotion
DeepMotion is a web-based AI motion capture and 3D animation platform with Animate 3D for video-to-animation and SayMotion for text-to-animation. It helps creators and teams generate motion in a browser and export results in common production formats.
Atlasly
Atlasly is an AI site analysis and feasibility tool for architects, developers, and planners, assessing constraints, walkability, and urban context in web app or MCP-connected AI assistants.
模袋云
模袋云 is an accessible online villa modeling software.
Vizcom
Vizcom is an AI design platform for product designers and teams that turns sketches into photorealistic 3D renders, with iteration and team workflows.
Amara
Amara is an AI-powered 3D creation tool focused on mesh generation and plugin-based actions. It offers a free entry point, tiered plans for individuals and teams, and staged early access.
Sloyd
Sloyd is a browser-based AI 3D model generator for text, images, and editable templates, built for game design, 3D printing, and common export workflows.