The agent that grows with you

Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

DFlash: Block Diffusion for Flash Speculative Decoding

A streamlined and customizable framework for efficient large model (LLM, VLM, AIGC) evaluation and performance benchmarking.

[NeurIPS 2025] Official Repo of Omni-R1: Reinforcement Learning for Omnimodal Reasoning via Two-System Collaboration

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

cuTile is a programming model for writing parallel kernels for NVIDIA GPUs

Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models

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

STEP-GUI: The top GUI agent solution in the galaxy. Developed by the StepFun-GELab team and powered by StepFun’s cutting-edge research capabilities.

A framework for efficient model inference with omni-modality models

Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

All-in-One Sandbox for AI Agents that combines Browser, Shell, File, MCP and VSCode Server in a single Docker container.

Tongyi Deep Research, the Leading Open-source Deep Research Agent

Checkpoint-engine is a simple middleware to update model weights in LLM inference engines

Code repo for "CritiPrefill: A Segment-wise Criticality-based Approach for Prefilling Acceleration in LLMs".

Empowering everyone to build reliable and efficient software.

Qwen-Image is a powerful image generation foundation model capable of complex text rendering and precise image editing.

Train speculative decoding models effortlessly and port them smoothly to SGLang serving.

vLLM’s reference system for K8S-native cluster-wide deployment with community-driven performance optimization

Supercharge Your LLM with the Fastest KV Cache Layer

Efficient Triton Kernels for LLM Training

Fast and memory-efficient exact attention

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

Spec-Bench: A Comprehensive Benchmark and Unified Evaluation Platform for Speculative Decoding (ACL 2024 Findings)

Muon is Scalable for LLM Training

Kimi K2 is the large language model series developed by Moonshot AI team