A curated list of awesome agentic operating systems, AI agents, and tools for the future of autonomous computing.
Inspired by the AgentOS list, this repository categorizes tools and frameworks designed to build, deploy, and manage AI agents and their ecosystems.
- Agent Frameworks & Runtimes
- Web Agents & Browsing
- Desktop OS & Cloud Platforms
- Visual & Creative AI
- Audio & Voice AI
- Developer Tools & Infrastructure
- Productivity & Personal Assistants
- Data, OCR & Knowledge Graphs
- Research & Specialized Agents
- block/goose - On-machine AI agent that automates development tasks from start to finish with MCP support.
- elizaOS/eliza - Multi-agent simulation framework with Discord, Telegram, and Twitter integration for autonomous agents.
- n8n-io/n8n - Fair-code workflow automation platform with native AI capabilities and 400+ integrations.
- simstudioai/sim - Open-source platform to build and deploy AI agent workflows.
- FlowiseAI/Flowise - Drag-and-drop interface for building customized LLM orchestration flows and AI agents.
- langflow-ai/langflow - Low-code platform for building and deploying AI-powered agents and workflows.
- SmythOS/sre - Cloud-native runtime for building, running, and managing intelligent agentic AI.
- activepieces/activepieces - Open-source AI automation framework with extensive MCP server support.
- FoundationAgents/OpenManus - An open-source implementation of an autonomous AI agent.
- letta-ai/letta - Platform for building stateful agents with advanced memory that learn over time.
- Significant-Gravitas/AutoGPT - An experimental open-source application showcasing the capabilities of the GPT-4 language model.
- open-interpreter/open-interpreter - Open-source AI agent that can execute code on your computer to perform tasks.
- huginn/huginn - System for creating agents that monitor and act on your behalf across the web.
- simular-ai/Agent-S - Open agentic framework designed to use computers just like a human.
- browseros-ai/BrowserOS - Open-source agentic browser serving as an alternative to proprietary AI browsing tools.
- browser-use/web-ui - Web interface for running and managing AI agents directly in your browser.
- nanobrowser/nanobrowser - Open-source Chrome extension for AI-powered web automation with multi-agent workflows.
- esinecan/agentic-ai-browser - AI-driven web automation agent utilizing Playwright for intelligent decision-making.
- unclecode/crawl4ai - Open-source, LLM-friendly web crawler and scraper for AI data gathering.
- MinorJerry/WebVoyager - End-to-end web agent framework powered by large multimodal models.
- browserless/browserless - Headless browser deployment platform optimized for Docker and cloud environments.
- monteslu/vibe-eyes - MCP server that enables LLMs to see and interact with browser-based applications.
- HeyPuter/puter - The Internet Computer - free, open-source, and self-hostable cloud desktop operating system.
- invoke-ai/InvokeAI - Leading creative engine for Stable Diffusion models to generate professional visual media.
- krea-ai/realtime-video - Open-source model for high-quality, realtime AI video generation.
- hpcaitech/Open-Sora - Democratizing efficient video production through open-source video generation models.
- MamtaRajpurohit/Cutalyst - Automated video editing AI that cuts, syncs, and subtitles with one click.
- gyoridavid/short-video-maker - Tool for creating short-form social media videos using MCP and REST APIs.
- shrimbly/node-banana - Free and open-source node-based generative workflow platform.
- TripoSR - Fast 3D object reconstruction from a single image using state-of-the-art AI.
- Sanster/IOPaint - AI-powered image inpainting tool for removing or replacing objects in photos.
- XingangPan/DragGAN - Interactive point-based manipulation for precise control over generative images.
- SkyworkAI/SkyReels-V2 - Generative model designed for creating infinite-length AI films.
- NVlabs/Sana - Efficient high-resolution image synthesis using Linear Diffusion Transformers.
- Tencent-Hunyuan/HunyuanVideo-1.5 - Leading lightweight video generation model for high-quality output.
- duixcom/Duix-Avatar - Open-source toolkit for AI avatar creation and digital human cloning.
- imlixinyang/FlashWorld - High-quality 3D scene generation framework that works within seconds.
- OHF-Voice/piper1-gpl - Fast and local neural text-to-speech engine for low-latency applications.
- Liquid4All/liquid-audio - Advanced speech-to-speech audio models developed by Liquid AI.
- RVC-Boss/GPT-SoVITS - Powerful few-shot voice cloning and text-to-speech model training framework.
- multimodal-art-projection/YuE - Open full-song music generation foundation model similar to Suno.ai.
- resemble-ai/chatterbox - State-of-the-art open-source text-to-speech engine for realistic voices.
- microsoft/VibeVoice - Open-source frontier voice AI for advanced audio synthesis.
- nazdridoy/kokoro-tts - CLI-based text-to-speech tool utilizing the Kokoro model for multiple languages.
- FunAudioLLM/Fun-Audio-Chat - Large audio language model built for natural, low-latency voice interactions.
- BerriAI/litellm - Python SDK and proxy server to call 100+ LLM APIs in a unified OpenAI format.
- mudler/LocalAI - Self-hosted, local-first open-source alternative to OpenAI and Claude APIs.
- voideditor/void - Open-source AI-powered code editor designed for agentic development.
- HKUDS/DeepCode - Open agentic coding framework for paper-to-code and web development tasks.
- unslothai/unsloth - Ultra-fast fine-tuning and reinforcement learning framework for LLMs.
- Klavis-AI/klavis - MCP integration platform for reliable tool use by AI agents at scale.
- zilliztech/claude-context - Code search MCP that makes entire codebases accessible to AI agents.
- ludwig-ai/ludwig - Low-code framework for building custom LLMs and neural networks easily.
- vercel-labs/json-render - Tool for dynamically rendering AI-generated JSON data into user interfaces.
- bytedance/UI-TARS-desktop - Open-source multimodal AI agent stack for desktop automation.
- jingyaogong/minimind - Educational project for training small-parameter GPT models from scratch.
- firecrawl/open-lovable - Chat with AI to clone and recreate any website as a modern React app in seconds.
- danny-avila/LibreChat - Enhanced ChatGPT clone with Agents, MCP, multi-model support, and enterprise-ready features.
- moeru-ai/airi - Self-hosted AI companion and VTuber platform capable of voice chat, game playing, and real-time interaction.
- openclaw/openclaw - Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞
- dinoki-ai/osaurus - AI edge infrastructure for macOS. Run local or cloud models, share tools across apps via MCP.
- character-ai/Ovi - Experimental AI character interaction tool from the Character.ai team.
- qwersyk/Newelle - Newelle - Your Ultimate Virtual Assistant.
- huggingface/chat-ui - The open source codebase powering HuggingChat with multi-model support.
- bytebot-ai/bytebot - Self-hosted AI desktop agent that automates computer tasks via natural language.
- BasedHardware/omi - AI wearables designed for real-time transcription and speech processing.
- chatboxai/chatbox - Versatile AI client supporting multiple models for daily productivity.
- rnchg/APT - AI productivity tool featuring built-in local ChatGPT and privacy-focused models.
- khoj-ai/khoj - AI second brain for searching documents, the web, and building custom agents.
- eigent-ai/eigent - Open-source coworker desktop application for enhancing individual productivity.
- janhq/jan - Open-source alternative to ChatGPT that runs 100% offline on your machine.
- ygwyg/system - AI-powered tool for controlling your Mac remotely from any location.
- open-webui/open-webui - User-friendly, self-hosted web interface for interacting with various LLMs.
- SillyTavern/SillyTavern - Advanced LLM frontend designed for power users and roleplay.
- mindverse/Second-Me - Platform to train your AI self and amplify your digital presence.
- souls-of-waifu - Self-hosted AI companion with voice chat and interactive capabilities.
- CaviraOSS/PageLM - Community-driven education platform transforming study materials into interactive resources.
- allenai/olmocr - Toolkit for linearizing PDFs to prepare datasets for LLM training.
- PaddlePaddle/PaddleOCR - Comprehensive OCR toolkit supporting 100+ languages and complex layouts.
- datalab-to/chandra - Specialized OCR model for parsing complex tables, forms, and handwriting.
- docling-project/docling - Tool for converting various document formats into AI-ready structured data.
- bytedance/Dolphin - Document image parsing framework using heterogeneous anchor prompting.
- getzep/graphiti - Tool for building real-time knowledge graphs to power AI agent memory.
- pathwaycom/pathway - Python ETL framework for real-time analytics, stream processing, and RAG.
- myriade-ai/myriade - AI-native data platform for exploring and transforming data warehouses.
- MemTensor/MemOS - AI memory operating system for persistent skill storage in agent systems.
- yazinsai/srt-ai - Translate SRT files to any language, using AI magic ✨
- machinewrapped/llm-subtrans - Open-source project using LLMs to translate subtitles across formats.
- ngxson/smolvlm-realtime-webcam - Real-time webcam demo with SmolVLM and llama.cpp server.
- SkyworkAI/DeepResearchAgent - Hierarchical multi-agent system for deep research and complex task solving.
- KeygraphHQ/shannon - Autonomous AI hacker designed to find exploits in web applications.
- virattt/dexter - Autonomous AI agent specialized in deep financial research and analysis.
- HKUDS/LightRAG - Simple and fast framework for retrieval-augmented generation.
- vikhyat/moondream - Tiny vision-language model optimized for edge devices and efficiency.
- facebookresearch/sam-3d-body - Inference code and models for the Segment Anything Model in 3D.
- NVlabs/OmniVinci - Omni-modal LLM for joint understanding of vision, audio, and language.
- eugeneyan/open-llms - Curated list of open LLMs available for commercial and research use.
- x1xhlol/system-prompts-and-models-of-ai-tools - Collection of system prompts and models for various AI tools.
Contributions are welcome! Please read the contribution guidelines first.