KernelBench: Can LLMs Write GPU Kernels? - Benchmark + Toolkit with Torch -> CUDA (+ more DSLs)

You like pytorch? You like micrograd? You love tinygrad! ❤️

MapAnything: Universal Feed-Forward Metric 3D Reconstruction

A Datacenter Scale Distributed Inference Serving Framework

The TinyLlama project is an open endeavor to pretrain a 1.1B Llama model on 3 trillion tokens.

[ICLR 2026] LLM/VLM gaming agents and model evaluation through games.

Efficient Long-context Language Model Training by Core Attention Disaggregation

Terminal UI for NVIDIA Nsight Systems profiles — timeline viewer, kernel navigator, NVTX hierarchy

Autoresearch for GPU kernels. Give it any PyTorch model, go to sleep, wake up to optimized Triton kernels.

Building the Virtuous Cycle for AI-driven LLM Systems

VideoNIAH: A Flexible Synthetic Method for Benchmarking Video MLLMs

Open-source unified multimodal model

Official repository for the paper "Light Forcing: Accelerating Autoregressive Video Diffusion via Sparse Attention"

Official codebase for "Self Forcing: Bridging Training and Inference in Autoregressive Video Diffusion" (NeurIPS 2025 Spotlight)

[ICLR 2026] Official Repo for Rolling Forcing: Autoregressive Long Video Diffusion in Real Time

MAGI-1: Autoregressive Video Generation at Scale

💖🧸 Self hosted, you-owned Grok Companion, a container of souls of waifu, cyber livings to bring them into our worlds, wishing to achieve Neuro-sama's altitude. Capable of realtime voice chat, Minec…

Helios: Real Real-Time Long Video Generation Model

Nano vLLM

Scalable Minecraft multiplayer data collection engine

The world's best AI personal assistant for email. Open source app to help you reach inbox zero fast.

A PyTorch-native inference engine with hybrid cache acceleration and massive parallelism for DiTs.

A powerful Web video editing UI framework, designed to help Web applications quickly integrate professional-grade video editing features.

AI video agents framework for next-gen video interactions and workflows.

My Python scripts to make high-quality figures for publications in top AI conferences and journals.

Give git-like & traceable memory to OpenClaw and any coding agents. By https://memov.ai/ aka Entire CLI for every coding agents by MCP.

OpenTinker is an RL-as-a-Service infrastructure for foundation models