通义千问发布Qwen-AgentWorld，一款原生语言世界模型，可在单一模型中模拟MCP、搜索、终端、SWE、Web、OS、Android共7种智能体环境。环境建模即训练目标，非事后适配。该模型在AgentWorldBench上性能超越Claude Opus 4.8和GPT-5.4。研究分两条路径：一是构建环境模拟基础模型；二是探索世界模型增强智能体训练——可控Sim RL（以LWM为环境的智能体强化学习）优于真实环境训练，而LWM预热（预测环境的学习）即使不经任何智能体特定微调，也能将预测知识迁移至智能体任务。
📣📣 Meet Qwen-AgentWorld — a native language world model that simulates 7 agent environments (MCP, Search, Terminal, SWE, Web, OS, Android) within a single model. Environment modeling is the training objective from day one, not a post-hoc adaptation.
🤔 LLMs are trained to be better agents — better at acting in environments. But nobody has trained them to model the environments themselves.
🗺️ Our roadmap: investigate how language world modeling can push the boundaries of general agent capabilities, along two routes:
1️⃣ Build a foundation model for environment simulation — outperforming Claude Opus 4.8 and GPT-5.4 on AgentWorldBench
2️⃣ Investigate how world modeling enhances agent training:
🔬 Controllable Sim RL (agentic RL with LWM as environments) surpasses training in real environments
🧠 Learning to predict environments (LWM warm-up) makes agents stronger — remarkably, even without any agent-specific training, this predictive knowledge transfers to agentic tasks with zero fine-tuning
📑 Paper: https://arxiv.org/abs/2606.24597
📖 Blog: https://qwen.ai/blog?id=qwen-agentworld
💻 GitHub: https://github.com/QwenLM/Qwen-AgentWorld
🤗 HuggingFace: https://huggingface.co/collections/Qwen/qwen-agentworld
🧩 ModelScope: https://modelscope.cn/collections/Qwen/Qwen-AgentWorld
来源：Qwen · x.com