1"use strict";(globalThis.webpackChunkiai_website=globalThis.webpackChunkiai_website||[]).push([[296],{82312(e,n,t){t.r(n),t.d(n,{assets:()=>l,contentTitle:()=>a,default:()=>h,frontMatter:()=>o,metadata:()=>i,toc:()=>d});var i=t(54600),s=t(74848),r=t(28453);const o={title:"AWorld: The Agent Runtime for Self-Improvement",date:new Date("2025-07-07T00:00:00.000Z"),authors:["inclusionai"],tags:["Release","Community"],custom_edit_url:null},a=void 0,l={authorsImageUrls:[void 0]},d=[{value:"Table of Contents",id:"table-of-contents",level:2},{value:"News",id:"news",level:2},{value:"Introduction",id:"introduction",level:2},{value:"Runtime Key Features",id:"runtime-key-features",level:3},{value:"Self-Improvement with Diverse Runtimes",id:"self-improvement-with-diverse-runtimes",level:3},{value:"Demo of GAIA Agent-Runtime",id:"demo-of-gaia-agent-runtime",level:3},{value:"Installation",id:"installation",level:2},{value:"Quick Start",id:"quick-start",level:2},{value:"Architecture",id:"architecture",level:2},{value:"Forward",id:"forward",level:3},{value:"Backward",id:"backward",level:3},{value:"Demo",id:"demo",level:2},{value:"Contributing",id:"contributing",level:2},{value:"License",id:"license",level:2},{value:"Star History",id:"star-history",level:2}];function c(e){const n={a:"a",blockquote:"blockquote",code:"code",em:"em",h2:"h2",h3:"h3",img:"img",li:"li",p:"p",pre:"pre",strong:"strong",table:"table",tbody:"tbody",td:"td",th:"th",thead:"thead",tr:"tr",ul:"ul",...(0,r.R)(),...e.components};return(0,s.jsxs)(s.Fragment,{children:[(0,s.jsxs)(n.p,{children:[(0,s.jsx)(n.em,{children:"\"Self-awareness: the hardest problem isn't solving within limits, it's discovering the own limitations\""}),"\n",(0,s.jsx)(n.a,{href:"https://x.com/InclusionAI666",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/twitter/follow/AWorld_AI?style=social",alt:"Twitter Follow"})}),"\n",(0,s.jsx)(n.a,{href:"https://raw.githubusercontent.com/inclusionAI/AWorld/main/readme_assets/aworld_wechat_qr.jpg",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/WeChat-Add%20us-green?logo=wechat&logoColor=white",alt:"WeChat QR Code"})}),"\n",(0,s.jsx)(n.a,{href:"https://discord.gg/b4Asj2ynMw",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Discord-Join%20us-blue?logo=discord&logoColor=white",alt:"Discord"})}),"\n",(0,s.jsx)(n.a,{href:"https://opensource.org/licenses/MIT",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/License-MIT-yellow.svg",alt:"License: MIT"})}),"\n",(0,s.jsx)(n.a,{href:"https://deepwiki.com/inclusionAI/AWorld",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/DeepWiki-Explore-blueviolet?logo=wikipedia&logoColor=white",alt:"DeepWiki"})})]}),"\n",(0,s.jsx)(n.h2,{id:"table-of-contents",children:"Table of Contents"}),"\n",(0,s.jsxs)(n.ul,{children:["\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.a,{href:"#news",children:"News"})," \u2014 Latest updates and announcements."]}),"\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.a,{href:"#introduction",children:"Introduction"})," \u2014 Overview and purpose of the project."]}),"\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.a,{href:"#installation",children:"Installation"})," \u2014 Step-by-step setup instructions."]}),"\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.a,{href:"#quick-start",children:"Quick Start"})," \u2014 Get started with usage examples."]}),"\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.a,{href:"#architecture",children:"Architecture"})," \u2014 Explore the multi-agent system design."]}),"\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.a,{href:"#demo",children:"Demo"})," \u2014 See the project in action with demonstrations."]}),"\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.a,{href:"#contributing",children:"Contributing"})," \u2014 How to get involved and contribute."]}),"\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.a,{href:"#license",children:"License"})," \u2014 Project licensing details."]}),"\n"]}),"\n",(0,s.jsx)(n.h2,{id:"news",children:"News"}),"\n",(0,s.jsxs)(n.ul,{children:["\n",(0,s.jsxs)(n.li,{children:["\ud83e\udda4 [2025/07/07] AWorld, as a runtime, is now ready for agentic training. See ",(0,s.jsx)(n.a,{href:"#self-improvement-with-diverse-runtimes",children:"Self-Improvement section"})," for details. We have updated our score to 77.08 on the GAIA test. Learn how to construct a GAIA runtime in the ",(0,s.jsx)(n.a,{href:"#demo-of-gaia-agent-runtime",children:"Demo section"}),"."]}),"\n",(0,s.jsxs)(n.li,{children:["\ud83e\udda9 [2025/06/19] We have updated our score to 72.43 on the GAIA test. Additionally, we have introduced a new local running mode. See ",(0,s.jsx)(n.code,{children:"./README-local.md"})," for detailed instructions."]}),"\n",(0,s.jsxs)(n.li,{children:["\ud83d\udc33 [2025/05/22] For quick GAIA evaluation, MCP tools, AWorld, and models are now available in a single Docker image. See ",(0,s.jsx)("code",{children:"./README-docker.md"})," for instructions and ",(0,s.jsx)(n.a,{href:"https://www.youtube.com/watch?v=kkYWeVvJKrg",children:"youtube video"})," for demo."]}),"\n",(0,s.jsxs)(n.li,{children:["\ud83e\udd73 [2025/05/13] AWorld has updated its state management for browser use and enhanced the video processing MCP server, achieving a score of 77.58 on GAIA validation (Pass@1 = 61.8) and maintaining its position as the top-ranked open-source framework. Learn more: ",(0,s.jsx)(n.a,{href:"https://huggingface.co/spaces
1/gaia-benchmark/leaderboard",children:"GAIA leaderboard"})]}),"\n",(0,s.jsxs)(n.li,{children:["\u2728 [2025/04/23] AWorld ranks 3rd on GAIA benchmark (69.7 avg) with impressive Pass@1 = 58.8, 1st among open-source frameworks. Reproduce with ",(0,s.jsx)("code",{children:"python examples/gaia/run.py"})]}),"\n"]}),"\n",(0,s.jsx)(n.h2,{id:"introduction",children:"Introduction"}),"\n",(0,s.jsx)(n.p,{children:"AWorld (Agent World) is a multi-agent playground that enables agents to collaborate and self-improve. The framework supports a wide range of applications, including but not limited to product prototype verification, foundation model training and Multi-Agent System (MAS) design meta-learning."}),"\n",(0,s.jsx)(n.h3,{id:"runtime-key-features",children:"Runtime Key Features"}),"\n",(0,s.jsxs)(n.table,{children:[(0,s.jsx)(n.thead,{children:(0,s.jsxs)(n.tr,{children:[(0,s.jsx)(n.th,{children:"1. Agent Construction"}),(0,s.jsx)(n.th,{children:"2. Topology Orchestration"}),(0,s.jsx)(n.th,{children:"3. Environments"})]})}),(0,s.jsx)(n.tbody,{children:(0,s.jsxs)(n.tr,{children:[(0,s.jsxs)(n.td,{children:["\u2022 \u2705 Support for various model services ",(0,s.jsx)("br",{})," \u2022 \u2705 Integration with MCP tools ",(0,s.jsx)("br",{})," \u2022 \u2705 Custom tool support"]}),(0,s.jsxs)(n.td,{children:["\u2022 \u2705 Protocol encapsulation between models and tools ",(0,s.jsx)("br",{})," \u2022 \u2705 Protocol encapsulation among agents"]}),(0,s.jsxs)(n.td,{children:["\u2022 \u2705 Runtime state management ",(0,s.jsx)("br",{})," \u2022 \u2705 State tracing support ",(0,s.jsx)("br",{})," \u2022 \u2705 Distributed, high-concurrency environments for training"]})]})})]}),"\n",(0,s.jsx)(n.h3,{id:"self-improvement-with-diverse-runtimes",children:"Self-Improvement with Diverse Runtimes"}),"\n",(0,s.jsx)(n.p,{children:"By constructing diverse runtime environments (with tools, agents, or models in them), AWorld aims to find the limitations of a model and push intelligence forward. Here we will record some of our work to prove the effectiveness of our proposal."}),"\n",(0,s.jsxs)(n.table,{children:[(0,s.jsx)(n.thead,{children:(0,s.jsxs)(n.tr,{children:[(0,s.jsx)(n.th,{children:"Category"}),(0,s.jsx)(n.th,{children:"Runtime"}),(0,s.jsx)(n.th,{children:"Performance"}),(0,s.jsx)(n.th,{children:"Key Information"})]})}),(0,s.jsxs)(n.tbody,{children:[(0,s.jsxs)(n.tr,{children:[(0,s.jsx)(n.td,{children:"Tool Use"}),(0,s.jsx)(n.td,{children:"Function call runtime to be released"}),(0,s.jsxs)(n.td,{children:["Competitive on BFCL benchmark ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.img,{src:"https://github.com/inclusionAI/AWorld/raw/main/readme_assets/funReason_BFCL.png",alt:"Agent Framework"})]}),(0,s.jsxs)(n.td,{children:[(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Dataset-Coming%20Soon-007ACC?style=for-the-badge&logo=dataset&logoColor=white",alt:"Dataset"})," ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.a,{href:"https://huggingface.co/Bingguang/FunReason",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Model-Hugging%20Face-FF6B6B?style=for-the-badge&logo=huggingface&logoColor=white",alt:"Model"})})," ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.a,{href:"https://arxiv.org/pdf/2505.20192",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Paper-arXiv-B31B1B?style=for-the-badge&logo=arxiv&logoColor=white",alt:"Paper"})})," ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Blog-Coming%20Soon-FF5722?style=for-the-badge&logo=blogger&logoColor=white",alt:"Blog"})," ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.a,{href:"https://github.com/BingguangHao/FunReason",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Code-GitHub-181717?style=for-the-badge&logo=github&logoColor=white",alt:"Code"})})]})]}),(0,s.jsxs)(n.tr,{children:[(0,s.jsx)(n.td,{children:"Deep Search"}),(0,s.jsx)(n.td,{children:"Search runtime to be released"}),(0,s.jsxs)(n.td,{children:["SOTA on HotpotQA benchmark ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.img,{src:"https://github.com/inclusionAI/AWorld/raw/main/readme_assets/HotpotQA_chart.png",alt:"Agent Framework"})]}),(0,s.jsxs)(n.td,{children:[(0,s.jsx)(n.a,{href:"https://github.com/inclusionAI/AgenticLearning",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Dataset-GitHub-181717?style=for-the-badge&logo=github&logoColor=white",alt:"Dataset"})})," ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.a,{href:"https://huggingface.co/collections/endertzw/rag-r1-68481d7694b3fca8b809aa29",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Model-Hugging%20Face-FF6B6B?style=for-the-badge&logo=huggingface&logoColor=white",alt:"Model"})})," ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.a,{href:"https://arxiv.org/abs/2507.02962",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Paper-arXiv-B31B1B?style=for-the-badge&logo=arxiv&logoColor=white",alt:"Paper"})})," ",(0,s.jsx)("br",{})," ",(0,s.jsx)(n.a,{href:"https://github.com/inclusionAI/AgenticLearning",children:(0,s.jsx)(n.img,{src:"https://img.shields.io/badge/Code-GitHub-181717?style=for-the-badge&logo=github&logoColor=white",alt:"Code"})})]})]})]})]}),"\n",(0,s.jsx)(n.h3,{id:"demo-of-gaia-agent-runtime",children:"Demo of GAIA Agent-Runtime"}),"\n",(0,s.jsx)(n.p,{children:(0,s.jsx)(n.img,{src:"https://github.com/inclusionAI/AWorld/raw/main/readme_assets/gaia_demo.gif",alt:"GAIA Agent Runtime Demo"})}),"\n",(0,s.jsxs)(n.p,{children:["Here we first introduce the ",(0,s.jsx)(n.strong,{children:"GAIA runtime"}),", which can be constructed on your local computer. It can be used for:"]}),"\n",(0,s.jsxs)(n.ul,{children:["\n",(0,s.jsx)(n.li,{children:(0,s.jsx)(n.strong,{children:"Product prototype verification"})}),"\n",(0,s.jsxs)(n.li,{children:[(0,s.jsx)(n.strong,{children:"Self-improvement training"})," (See ",(0,s.jsx)(n.a,{href:"#backward",children:"training pipeline"})," for details)"]}),"\n"]}),"\n",(0,s.jsxs)(n.p,{children:["Follow the instructions in ",(0,s.jsx)(n.a,{href:"https://github.com/inclusionAI/AWorld/blob/main/examples/gaia/README.md",children:(0,s.jsx)(n.code,{children:"./examples/gaia/README.md"})})," to initialize the GAIA agent runtime and run the demo shown above."]}),"\n",(0,s.jsxs)(n.blockquote,{children:["\n",(0,s.jsx)(n.p,{children:(0,s.jsx)(n.strong,{children:"Want to build your own multi-agent system? Check out the detailed tutorials below to get started! \u2b07\ufe0f\u2b07\ufe0f\u2b07\ufe0f"})}),"\n"]}),"\n",(0,s.jsx)(n.h2,{id:"installation",children:"Installation"}),"\n",(0,s.jsx)(n.p,{children:"Python>=3.11:"}),"\n",(0,s.jsx)(n.pre,{children:(0,s.jsx)(n.code,{className:"language-bash",children:"git clone https://github.com/inclusionAI/AWorld\ncd AWorld\npython setup.py install\n"})}
1),"\n",(0,s.jsx)(n.h2,{id:"quick-start",children:"Quick Start"}),"\n",(0,s.jsxs)(n.blockquote,{children:["\n",(0,s.jsx)(n.p,{children:"Here's a quick start guide to: (1) create your first agent; (2) equip it with a MCP tool; (3) assign a teammate; and (4) answer a user query through teamwork."}),"\n"]}),"\n",(0,s.jsx)(n.pre,{children:(0,s.jsx)(n.code,{className:"language-python",children:'from aworld.config.conf import AgentConfig\nfrom aworld.agents.llm_agent import Agent\nfrom aworld.runner import Runners\nfrom aworld.core.agent.swarm import Swarm\n\nif __name__ == \'__main__\':\n agent_config = AgentConfig(\n llm_provider="openai",\n llm_model_name="gpt-4o",\n\n # Set via environment variable or direct configuration\n # llm_api_key="YOUR_API_KEY",\n # llm_base_url="https://api.openai.com/v1"\n )\n\n # Register the MCP tool here, or create a separate configuration file.\n mcp_config = {\n "mcpServers": {\n "amap-amap-sse": {\n "type": "sse",\n "url": "https://mcp.amap.com/sse?key=YOUR_API_KEY",\n "timeout": 5,\n "sse_read_timeout": 300\n }\n }\n }\n\n # Create your first agent equipped with an MCP tool\n search = Agent(\n conf=agent_config,\n name="search_agent",\n system_prompt="You are a helpful agent.",\n mcp_servers=["amap-amap-sse"], # MCP server name for agent to use\n mcp_config=mcp_config\n )\n\n # Add a new teammate to the agent\n summary = Agent(\n conf=agent_config,\n name="summary_agent",\n system_prompt="You are a helpful summary agent."\n )\n\n # Collaborate as a team; the default is a static workflow\n swarm = Swarm(search, summary)\n\n # Run agent team\n res = Runners.sync_run(input="Hotels within 1 kilometer of West Lake in Hangzhou",\n swarm=swarm)\n print(res)\n'})}),"\n",(0,s.jsx)(n.h2,{id:"architecture",children:"Architecture"}),"\n",(0,s.jsx)(n.p,{children:"AWorld is designed to achieve two primary objectives: (1) provide an efficient forward process, and (2) facilitate diverse backward processes, including but not limited to foundation model training and system design meta-learning."}),"\n",(0,s.jsx)(n.h3,{id:"forward",children:"Forward"}),"\n",(0,s.jsxs)(n.blockquote,{children:["\n",(0,s.jsx)(n.p,{children:"An illustration of the runtime, showing the message workflow when Agent1 receives a query from a user."}),"\n"]}),"\n",(0,s.jsx)(n.p,{children:(0,s.jsx)(n.img,{src:"https://github.com/inclusionAI/AWorld/raw/main/readme_assets/runtime.jpg",alt:""})}),"\n",(0,s.jsx)(n.h3,{id:"backward",children:"Backward"}),"\n",(0,s.jsxs)(n.blockquote,{children:["\n",(0,s.jsx)(n.p,{children:"During training, an action-state rollout demonstration using AWorld's distributed environments."}),"\n"]}),"\n",(0,s.jsx)(n.p,{children:(0,s.jsx)(n.img,{src:"https://github.com/inclusionAI/AWorld/raw/main/readme_assets/agent_training2.jpg",alt:""})}),"\n",(0,s.jsx)(n.h2,{id:"demo",children:"Demo"}),"\n",(0,s.jsxs)(n.blockquote,{children:["\n",(0,s.jsxs)(n.p,{children:["Running Pre-defined Agents (e.g., see ",(0,s.jsx)(n.a,{href:"https://github.com/inclusionAI/AWorld/blob/main/examples/browsers/run.py",children:"demo code"}),"). Below are demonstration videos showcasing AWorld's capabilities across various agent configurations and environments."]}),"\n"]}),"\n",(0,s.jsxs)("table",{children:[(0,s.jsxs)("tr",{children:[(0,s.jsx)("th",{children:"Mode"}),(0,s.jsx)("th",{children:"Type"}),(0,s.jsx)("th",{children:"Demo"})]}),(0,s.jsxs)("tr",{children:[(0,s.jsx)("td",{rowspan:"2",children:"Single Agent"}),(0,s.jsx)("td",{children:"Browser use"}),(0,s.jsx)("td",{children:(0,s.jsxs)("a",{href:"https://www.youtube.com/watch?v=R7keOLrRDoM",target:"_blank",children:[(0,s.jsx)("img",{src:"https://img.youtube.com/vi/R7keOLrRDoM/0.jpg",alt:"AWorld Browser Demo on YouTube",width:"95%"}),(0,s.jsx)("br",{}),(0,s.jsx)("p",{align:"center",children:"\u25b6\ufe0f Watch Browser Demo on YouTube"})]})})]}),(0,s.jsxs)("tr",{children:[(0,s.jsx)("td",{children:"Phone use"}),(0,s.jsx)("td",{children:(0,s.jsxs)("a",{href:"https://www.youtube.com/watch?v=TYh3iqDeIoQ",target:"_blank",children:[(0,s.jsx)("img",{src:"https://img.youtube.com/vi/TYh3iqDeIoQ/0.jpg",alt:"AWorld Mobile Demo on YouTube",width:"95%"}),(0,s.jsx)("br",{}),(0,s.jsx)("p",{align:"center",children:"\u25b6\ufe0f Watch Mobile Demo on YouTube"})]})})]}),(0,s.jsxs)("tr",{children:[(0,s.jsx)("td",{rowspan:"3",children:"Multi Agent"}),(0,s.jsx)("td",{children:"Cooperative Teams"}),(0,s.jsx)("td",{children:(0,s.jsxs)("a",{href:"https://www.youtube.com/watch?v=sEsgasRrlTs",target:"_blank",children:[(0,s.jsx)("img",{src:"https://img.youtube.com/vi/sEsgasRrlTs/0.jpg",alt:"AWorld Travel Demo on YouTube",width:"95%"}),(0,s.jsx)("br",{}),(0,s.jsx)("p",{align:"center",children:"\u25b6\ufe0f Watch Travel Demo on YouTube"})]})})]}),(0,s.jsxs)("tr",{children:[(0,s.jsx)("td",{children:"Competitive Teams"}),(0,s.jsx)("td",{children:(0,s.jsxs)("a",{href:"https://www.youtube.com/watch?v=_CPdhoP4YTg",target:"_blank",children:[(0,s.jsx)("img",{src:"https://img.youtube.com/vi/_CPdhoP4YTg/0.jpg",alt:"AWorld Debate Demo on YouTube",width:"95%"}),(0,s.jsx)("br",{}),(0,s.jsx)("p",{align:"center",children:"\u25b6\ufe0f Watch Debate Arena on YouTube"})]})})]}),(0,s.jsxs)("tr",{children:[(0,s.jsx)("td",{children:"Mixed of both Teams"}),(0,s.jsxs)("td",{align:"center",children:[(0,s.jsx)("i",{children:"Coming Soon"})," \ud83d\ude80"]})]})]}),"\n",(0,s.jsx)(n.h2,{id:"contributing",children:"Contributing"}),"\n",(0,s.jsx)(n.p,{children:"We warmly welcome developers to join us in building and improving AWorld! Whether you're interested in enhancing the framework, fixing bugs, or adding new features, your contributions are valuable to us."}),"\n",(0,s.jsx)(n.p,{children:"For academic citations or wish to contact us, please use the following BibTeX entry:"}),"\n",(0,s.jsx)(n.pre,{children:(0,s.jsx)(n.code,{className:"language-bibtex",children:"@software{aworld2025,\n author = {Agent Team at inclusionAI},\n title = {AWorld: Enabling Agent Self-Improvement through Interactive Experience with Dynamic Runtime},\n year = {2025},\n url = {https://github.com/inclusionAI/AWorld},\n version = {0.1.0},\n publisher = {GitHub},\n email = {chenyi.zcy at antgroup.com}\n}\n"})}),"\n",(0,s.jsx)(n.h2,{id:"license",children:"License"}),"\n",(0,s.jsxs)(n.p,{children:["This project is licensed under the MIT License - see the ",(0,s.jsx)(n.a,{href:"https://github.com/inclusionAI/AWorld/blob/main/LICENSE",children:"LICENSE"})," file for details."]}),"\n",(0,s.jsx)(n.h2,{id:"star-history",children:"Star History"}),"\n",(0,s.jsx)(n.p,{children:(0,s.jsx)(n.img,{src:"https://api.star-history.com/svg?repos=inclusionAI/AWorld&type=Date",alt:""})})]})}function h(e={}){const{wrapper:n}={...(0,r.R)(),...e.components};return n?(0,s.jsx)(n,{...e,children:(0,s.jsx)(c,{...e})}):c(e)}},28453(e,n,t){t.d(n,{R:()=>o,x:()=>a});var i=t(96540);const s={},r=i.createContext(s);function o(e){const n=i.useContext(r);return i.useMemo(function(){return"function"==typeof e?e(n):{...n,...e}},[n,e])}function a(e){let n;return n=e.disableParentContext?"function"==typeof e.components?e.components(s):e.components||s:o(e.components),i.createElement(r.Provider,{value:n},e.children)}},54600(e){e.exports=JSON.parse('{"permalink":"/blog/aworld","source":"@site/blog/aworld/index.mdx","title":"AWorld: The Agent Runtime for Self-Improvement","description":"\\"Self-awareness: the hardest problem isn\'t solving within limits, it\'s discovering the own limitations\\"","date":"2025-07-07T00:00:00.000Z","tags":[{"inline":true,"label":"Release","permalink":"/blog/tags/release"},{"inline":true,"label":"Community","permalink":"/blog/tags/community"}],"readingTime":7.48,"hasTruncateMarker":false,"authors":[{"name":"inclusionAI","title":"Ant Group","url":"https://github.com/inclusionAI","email":"[email protected]","imageURL":"https://avatars.githubusercontent.com/u/199075982?s=200&v=4","key":"inclusionai","page":null}],"frontMatter":{"title":"AWorld: The Agent Runtime for Self-Improvement","date":"2025-07-07T00:00:00.000Z","authors":["inclusionai"],"tags":["Release","Community"],"custom_edit_url":null},"unlisted":false,"prevItem":{"title":"ABench: An Evolving Open-Source Benchmark","permalink":"/blog/abench"},"nextItem":{"title":"Open Source LLM Development 2025: Landscape, Trends and Insights","permalink":"/blog/llm-landscape-2025"}}')}}]);
Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.