Official implementation for paper "How Far Are We from Genuinely Useful Deep Research Agents?"

DiffusionNFT: Online Diffusion Reinforcement with Forward Process

OmniVinci is an omni-modal LLM for joint understanding of vision, audio, and language.

NEO Series: Native Vision-Language Models from First Principles

Open-source implementation of AlphaEvolve

SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?

Data and code for FreshLLMs (https://arxiv.org/abs/2310.03214)

The official repo of "WebExplorer: Explore and Evolve for Training Long-Horizon Web Agents"

Code for paper: Reinforced Vision Perception with Tools

Python binding for curl-impersonate fork via cffi. A http client that can impersonate browser tls/ja3/http2 fingerprints.

MixGRPO: Unlocking Flow-based GRPO Efficiency with Mixed ODE-SDE

TempFlow-GRPO (Temporal Flow GRPO), a principled GRPO framework that captures and exploits the temporal structure inherent in flow-based generation.

The 100 line AI agent that solves GitHub issues or helps you in your command line. Radically simple, no huge configs, no giant monorepo—but scores >74% on SWE-bench verified!

Multimodal Referring Segmentation

Think Beyond Images

The official github repo for "Diffusion Language Models are Super Data Learners".

ACTIVE-O3: Empowering Multimodal Large Language Models with Active Perception via GRPO

Codebase for paper ToolVQA: A Dataset for Multi-step Reasoning VQA with External Tools

[NIPS 25'] Evaluation code of paper "KRIS-Bench: Benchmarking Next-Level Intelligent Image Editing Models"

Hierarchical Reasoning Model Official Release

A Scientific Multimodal Foundation Model

Scientists' First Exam: Probing Cognitive Abilities of MLLM via Perception, Understanding, and Reasoning