GPU-accelerated rigid-body dynamics solver for robot learning
Professional-grade creative suite for game developers and creators, running entirely in the browser
Stage
Not given
MVP
Founded
Not given
2025
Based in
Not given
CH
Pricing model
Open source
Freemium
Pricing
Not given
Free: Inkwell, Vectra, and PDF Studio. Pro: $5/month or $50/year (all modules). One-time licenses available: Full suite for 1 year, Lifetime module license, Lifetime license for all modules. aukimi Desktop: $119 one-time payment.
Key features
3.8× PhysX throughput on 20 benchmarks geometric mean
13.0× performance on heterogeneous task mixtures
Per-environment work scheduling for efficiency
Reduced-coordinate articulation dynamics
Differentiable simulation
Rigid-deformable coupling support
Compliant and hydroelastic contact
62% less GPU memory than PhysX at matched batch size
3.90 million environment steps per second peak throughput
14 integrated modules for 2D, 3D, animation, video, audio, and game development
Real-time synchronization across modules
Support for 59+ file formats
Cloud auto-save and sync
Cross-platform: desktop, tablet, and mobile
Pen and tablet support with pressure sensitivity
Local project storage with optional cloud sync
Asset Store marketplace for 2D art, 3D models, and templates
AI features designed to assist without replacing creativity
Available in 28 languages including right-to-left layouts
What makes it different
Designed for heterogeneous workloads with different robots and tasks simulated together efficiently
Integrated into open-source Newton for seamless adoption without separate physics stack
Supports diverse physical interactions beyond rigid bodies
Enables scaling of diverse robot experience for general-purpose learning
Complete unified creative workflow in one browser-based application, eliminating the need for multiple subscriptions
No forced subscriptions with lifetime module purchase options
Built on proven open-source engines with studio-grade performance
User data and creations remain under creator control with no AI training on user work