Last updated 08-13-2026
Category:
Reviews:
Join thousands of AI enthusiasts in the World of AI!
Neuralangelo Research Reconstructs 3D Scenes | NVIDIA
Neuralangelo is an NVIDIA Research model that reconstructs detailed 3D objects and scenes from ordinary 2D video. Feed clips filmed from multiple angles and the system builds textured meshes of buildings, sculptures, vehicles, and interiors suitable for games, digital twins, and robotics sims.
Earlier neural reconstruction methods blurred fine textures like roof shingles, glass, and marble. Neuralangelo extends instant neural graphics primitives from NVIDIA Instant NeRF to capture those surfaces with higher fidelity, so smartphone footage can produce assets creative teams can import into design tools without a laser scan or manual modeling pass.
Researchers, game artists, and robotics engineers use Neuralangelo to digitize real-world objects from Michelangelo-scale statues to campus buildings. The project is open source on GitHub under nvlabs/neuralangelo with no commercial license fees for the research code itself.
Reconstructs 3D objects and large scenes from 2D video filmed at multiple angles
Captures fine textures like shingles, glass, and marble better than prior neural methods
Built on instant neural graphics primitives from NVIDIA Instant NeRF technology
Exports detailed meshes for game engines, VR, digital twins, and robotics pipelines
Open-source code published at github.com/nvlabs/neuralangelo
Demonstrated on statues, vehicles, and full building interiors from phone footage
Presented at CVPR 2023 alongside nearly 30 other NVIDIA Research projects
High-fidelity textures on shingles, glass, and marble from phone video.
Open-source release on GitHub under NVIDIA Research.
Outputs usable meshes for games, VR, and robotics pipelines.
Handles both small objects and large building-scale scenes.
Research codebase requires GPU setup and technical expertise to run.
Needs multi-angle video input rather than a single photo.
No hosted SaaS; you run Neuralangelo locally or on your own infrastructure.
What is Neuralangelo?
Neuralangelo is an NVIDIA Research AI model that reconstructs 3D structures from 2D video clips. It generates textured virtual replicas of real-world objects and scenes for creative and robotics workflows.
Is Neuralangelo free to use?
Yes. Neuralangelo is available as open-source research code on GitHub at nvlabs/neuralangelo. NVIDIA publishes the model for developers and researchers to download and run.
What input does Neuralangelo need?
Neuralangelo takes 2D video of an object or scene captured from multiple viewpoints. The model selects frames, estimates camera positions, and optimizes a detailed 3D representation.
What can I build with Neuralangelo output?
Neuralangelo meshes can be imported into design apps for video games, virtual reality experiences, industrial digital twins, and robotics simulation environments.
How is Neuralangelo different from Instant NeRF?
Neuralangelo builds on instant neural graphics primitives from NVIDIA Instant NeRF but focuses on sharper surface textures and large-scale scene reconstruction from casual video.
Where is the Neuralangelo code hosted?
Neuralangelo source code is on GitHub at github.com/nvlabs/neuralangelo. NVIDIA announced public availability in an editor note on its official blog post.
Who developed Neuralangelo?
Neuralangelo was developed by NVIDIA Research, including senior director Ming-Yu Liu and collaborators, and was presented at the CVPR 2023 conference.