Large Action Model framework to develop AI Web Agents

[NeurIPS 2024] OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments

Open-source adapter between multimodal models and desktop/web GUIs. Record a workflow once, then compile it into a deterministic, self-healing replay (openadapt-flow) — or condition, train, and evaluate agents on it. Modular meta-package; install only the extras you need.

Hercules is the world’s first open-source testing agent, enabling UI, API, Security, Accessibility, and Visual validations – all without code or maintenance. Automate testing effortlessly and let Hercules handle the heavy lifting! ⚡

Awesome Large Action Model (LAM): Models that could help gets things done.

OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks

Official repository of the spotlight ICML 2025 paper, PokeChamp: an Expert-level Minimax Language Agent.

[NeurIPS 2025 Spotlight] Scaling Computer-Use Grounding via UI Decomposition and Synthesis

An open source alternative to the Rabbit R1 directly on your phone

Playback Electron App

an automation pipeline that discovers, extracts, and validates e-commerce coupon codes using headless browsing with LAMs

R-LAM is a reproducibility-constrained execution framework for Large Action Models in scientific workflow automation. It enables adaptive, agent-driven workflow execution while enforcing strict guarantees on auditability, determinism, and replayability.

Rabbit (R1) — Large Action Model handheld AI device + LAM Playground

Agentic AI System with local PC MCP Execution in help with remote server's agentic AI process with remote MCP tools and Agents.

Experimental evaluation of Large Action Model execution paradigms for reproducible scientific workflows

Fine-tuning mistral model to accurately time stamp relevant youtube video segment, later labelled using trained IDM.

The first open-source Artificial Narrow Intelligence generalist agentic framework Computer-Using-Agent that fully operates graphical-user-interfaces (GUIs) by using only natural language. Uses Visualization-of-Thought and Chain-of-Thought reasoning to elicit spatial reasoning and perception, emulates, plans and simulates synthetic HID interactions.

Generate AI images and videos with 200+ open-source models. Access state-of-the-art tools without filters, subscriptions, or ecosystem restrictions.

A high-performance AI workstation showcasing LLM, LCM, and LAM paradigms. Built with React 19, Tailwind CSS 4, and Google Gemini 2.5 Flash.

Control Android phones from your desktop or AI agent to tap, type, launch apps, inspect UI, and capture screenshots over LAN