[CVPR 2025 Best Paper Award] VGGT: Visual Geometry Grounded Transformer

[ICLR 2026] Streaming 4D Visual Geometry Transformer

VGGT-SLAM: Dense RGB SLAM Optimized on the SL(4) Manifold

Official implement of VGGT-Long

[ICLR 2026] FastVGGT: Fast Visual Geometry Transformer

Unified Codebase for Advanced World Models.

The official implementation of InfiniteVGGT

[CVPR 2026 Hightlight] OmniVGGT: Omni-Modality Driven Visual Geometry Grounded Transformer

VGGT-X: When VGGT Meets Dense Novel View Synthesis

BRIDGE BriGeS ChronoDepth Depth Any Video Depth Anything Depth Pro DepthCrafter Distill Any Depth Elastic3D Eye2Eye FE2E FlashDepth GeometryCrafter HairGuard M2SVid MegaSaM Metric3D MoGe MoRE NVDS Pixel-Perfect Depth Restereo SpatialTrackerV2 StableDPT StereoCrafter StereoPilot StereoWorld SVG Uni4D UniDepth VGGT Video Depth Anything π^3

The official implementation of the paper “VGGT4D: Mining Motion Cues in Visual Geometry Transformers for 4D Scene Reconstruction.”

Track4World: Feedforward World-centric Dense 3D Tracking of All Pixels

[CVPR 2026] LiteVGGT: Boosting Vanilla VGGT via Geometry-aware Cached Token Merging

[CVPR'26] Official implementation of "Emergent Outlier View Rejection in Visual Geometry Grounded Transformers"

Repository for running the VGGT model in PyTorch

This is a list of relevant papers for 3D Geometric Foundation Models and Applications.

Official code for CVPR 2026 paper: VGGT-Det: Mining VGGT Internal Priors for Sensor-Geometry-Free Multi-View Indoor 3D Object Detection

[ICLR 2026] Code for QuantVGGT: Quantized Visual Geometry Grounded Transformer

Code for Faster VGGT with Block-Sparse Global Attention

[ICLR 2026] PAGE-4D: Disentangled Pose and Geometry Estimation for VGGT-4D Perception

VGGT 3D Vision Agent optimized for Apple Silicon with Metal Performance Shaders

Official implementation of “4D LangVGGT: 4D Language-Visual Geometry Grounded Transformer”

Blender addon for vggt 3D reconstruction

VGGT optimized for VRAM usage - 6x less VRAM, 150 images in 8GB, 1100 images in 32GB

[CVPR 2026 Findings] SwiftVGGT: A Scalable Visual Geometry Grounded Transformer for Large-Scale Scenes

ONNX models of VGGT

Accelerate VGGT with efficient desciptor-based global attention

😎 A curated list of CVPR 2025 Oral paper. Total 96

[2026 RA-L] LiDAR VGGT: Cross-Modal Coarse-to-Fine Fusion for Globally Consistent and Metric-Scale Dense Mapping

Optimizing Monocular Depth Estimation with TensorRT: Model Conversion, Inference Acceleration, and 3D Reconstruction