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

Fuji is an AI agent that lives in your browser's sidepanel. You can now get tasks done online with a single command!

Code and data for "Lumos: Learning Agents with Unified Data, Modular Design, and Open-Source LLMs"

This is the repo for the paper "OS Agents: A Survey on MLLM-based Agents for Computer, Phone and Browser Use" (ACL 2025 Oral).

Official repository for "RLVR-World: Training World Models with Reinforcement Learning" (NeurIPS 2025), https://arxiv.org/abs/2505.13934

Web-Use is an agent for web browsing and scraping websites.

Run Surfer-H agents powered by Holo1 using the Surfer-H-CLI. Includes example tasks, scripts, and configurations.

[NAACL2025] LiteWebAgent: The Open-Source Suite for VLM-Based Web-Agent Applications

The Library for LLM-based multi-agent applications

Fathom-DeepResearch: Unlocking Long Horizon Information Retrieval And Synthesis For SLMs

Opensource benchmark evaluating web operators/agents performance

Agent Skill Induction: "Inducing Programmatic Skills for Agentic Tasks"

Code for 🌍 UI-Simulator: LLMs as Scalable, General-Purpose Simulators For Evolving Digital Agent Training

Navi-Bench: benchmarking web agents on everyday tasks directly on real websites

vibebin is an Incus/LXC-based platform for self-hosting persistent AI coding agent sandboxes with Caddy reverse proxy and direct SSH routing to containers (suitable for VS Code remote ssh). Create and host your vibe-coded apps on a single VPS/server.

Screen recording and computer interaction capture tool that records keyboard/mouse input, screen video, DOM snapshots, and accessibility trees. Perfect for creating datasets to train and evaluate computer-use AI models.

Screen recording and computer interaction capture tool that records keyboard/mouse input, screen video, DOM snapshots, and accessibility trees. Perfect for creating datasets to train and evaluate computer-use AI models.

Python scripts for generating and categorizing web browsing tasks for benchmark datasets

Agent-CE is a containerized continuous evaluation (CE) platform for web browsing agents. It provides production-ready Docker images and CI/CD pipelines for running and evaluating multiple agent frameworks including Browser Use, Notte, Anthropic Computer Use, and OpenAI Computer Use.

Neurosim is a Python framework for building, running, and evaluating AI agent systems. It provides core primitives for agent evaluation, cloud storage integration, and an LLM-as-a-judge system for automated scoring.