Cognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory across sessions with a self-hosted knowledge graph engine.

Mixture-of-experts (MoE) training megakernel for NVL72s

A framework for serving and evaluating LLM routers - save LLM costs without compromising quality

Library for specialized dense and sparse matrix operations, and deep learning primitives.

slime is an LLM post-training framework for RL Scaling.

Modular RDMA Interface

Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.

LMCache: Supercharge Your LLM with the Fastest KV Cache Layer

AI PC starter app for doing AI image creation, image stylizing, and chatbot on a PC powered by an Intel® Arc™ GPU.

Daytona is a Secure and Elastic Infrastructure for Running AI-Generated Code

Open Model Engine (OME) — Kubernetes operator for LLM serving, GPU scheduling, and model lifecycle management. Works with SGLang, vLLM, TensorRT-LLM, and Triton

NVIDIA Infra Controller - Hardware Lifecycle Management and multitenant networking

ToolOrchestra is an end-to-end RL training framework for orchestrating tools and agentic workflows.

TurboDiffusion: 100–200× Acceleration for Video Diffusion Models

Google 💚 MCP

LLaDA2.0 is the diffusion language model series developed by InclusionAI team, Ant Group.

Helpful kernel tutorials, examples and SKILLs for tile-based GPU programming

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.

Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.

LLM Council works together to answer your hardest questions

SGLang is a high-performance serving framework for large language models and multimodal models.

Simple framework for training and evaluating math reasoning agents using local models, GRPO and vLLM.

The official implementation of dLLM-Var

[NeurIPS'25] dKV-Cache: The Cache for Diffusion Language Models

A collection of tricks and tools to speed up transformer models

dLLM: Simple Diffusion Language Modeling

A toolkit for developing and comparing reinforcement learning algorithms.

Open diffusion language model for code generation — releasing pretraining, evaluation, inference, and checkpoints.

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

Example how to use SmolAgent with the diffusion based Inception Model