RLinf: Reinforcement Learning Infrastructure for Embodied and Agentic AI

Official Codebase for "Do as I Do: Dexterous Manipulation Data from Everyday Human Videos"

Official repository for "SO-101 Bench: Measuring the Gap Between Semantic and Geometric Competence in Robot Foundation Models"

A Curated List of Vision-Language-Action (VLA) and World Action Models (WAM) Research and Beyond

Cosmos-Predict2.5, the latest version of the Cosmos World Foundation Models (WFMs) family, specialized for simulating and predicting the future state of the world in the form of video.

X-MOBILITY

CUDA-accelerated visual odometry and SLAM library for fast, accurate, low-latency camera tracking

AIBuildAI Inc 

Hundreds of models & providers. One command to find what runs on your hardware.

[CVPR 2026] Fast-FoundationStereo: Real-Time Zero-Shot Stereo Matching

Vane is an AI-powered answering engine.

VLA model interpretability tools

Code to pretrain, fine-tune, and evaluate DreamZero and run sim & real-world evals

The AI that really does things. Any OS. Any Platform. The lobster way. 🦞

Unified Embedding Trainer for LLMs & VLMs: Fine-tune Qwen3, ColPali, & CLIP. Scalable from single 24GB consumer GPUs to multi-card clusters. Features Gradient Cache, Matryoshka (MRL) & LoRA for SOT…

📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG

Paper implementation

A beginner-friendly SLAM mini-course with Jupyter notebooks — covering Bayes Filters, Kalman Filters, Particle Filters, and Graph-based SLAM with hands-on Python examples.

A modular, first-principles walk-through of Vision-Language-Action models

The repository provides code for running inference and finetuning with the Meta Segment Anything Model 3 (SAM 3), links for downloading the trained model checkpoints, and example notebooks that sho…

Central repository for the Holoscan Ecosystem

Depth Anything 3

[ICML 2026] WorldMirror: Fast and Universal 3D reconstruction model for versatile tasks

An open-source, GPU-accelerated physics simulation engine built upon NVIDIA Warp, specifically targeting roboticists and simulation researchers.

[CVPR 2025 Best Paper Nomination] FoundationStereo: Zero-Shot Stereo Matching

Use YOLOv8 with a DJI Tello Drone! Controlled with the Tello app and provides video results post-flight.

A DroneBlocks course on drone programming with Tello using Python scripts

An open source, modular, lightweight, and extensible RAG application stack, designed for anyone and everyone.

Get your documents ready for gen AI

A unified library of SOTA model optimization techniques like quantization, distillation, pruning, neural architecture search, speculative decoding, etc. It compresses deep learning models for downs…