The Year in AI Papers
The year/Independent research

Paper 2512.24615

Youtu-Agent: Scaling Agent Productivity with Automated Generation and Hybrid Policy Optimization

Yuchen Shi, Yuzheng Cai, Siqi Cai, Zihan Xu, Lichao Chen, Yulei Qin, Zhijian Zhou, Xiang Fei, Chaofan Qiu, Xiaoyu Tan, Gang Li, Zongyi Li, et al.

Published
Dec 2025
Research lab
Independent
Citations
5
GitHub
4.6K stars

01 In brief

Summary

Youtu-Agent is a modular framework addressing high configuration costs and static capabilities in LLM agents.

It decouples environments, toolkits, and agents via a YAML-based system, enabling automated generation through a deterministic Workflow mode and a flexible Meta-Agent mode that synthesizes tools, prompts, and configurations.

For continuous improvement, it introduces Agent Practice, a training-free method using group relative optimization to accumulate in-context experience, and Agent RL, which integrates with distributed training frameworks for scalable and stable reinforcement learning.

Experiments show state-of-the-art results on WebWalkerQA (71.47%) and GAIA (72.8%) using open-weight models.

Automated tool synthesis achieves over 81% success rate.

Agent Practice improves AIME 2024/2025 by +2.7% and +5.4% with only 100 samples and $18 cost.

Agent RL training achieves 40% speedup, scales to 128 GPUs, and improves Qwen2.5-7B accuracy on AIME 2024 from 10% to 45%, with gains up to 35% on math and 21% on QA benchmarks.

The framework also includes Tip, an on-device desktop assistant for practical deployment.

02 From the paper

Abstract

Existing Large Language Model (LLM) agent frameworks face two significant challenges: high configuration costs and static capabilities. Building a high-quality agent often requires extensive manual effort in tool integration and prompt engineering, while deployed agents struggle to adapt to dynamic environments without expensive fine-tuning. To address these issues, we propose \textbf{Youtu-Agent}, a modular framework designed for the automated generation and continuous evolution of LLM agents. Youtu-Agent features a structured configuration system that decouples execution environments, toolkits, and context management, enabling flexible reuse and automated synthesis. We introduce two generation paradigms: a \textbf{Workflow} mode for standard tasks and a \textbf{Meta-Agent} mode for complex, non-standard requirements, capable of automatically generating tool code, prompts, and configurations. Furthermore, Youtu-Agent establishes a hybrid policy optimization system: (1) an \textbf{Agent Practice} module that enables agents to accumulate experience and improve performance through in-context optimization without parameter updates; and (2) an \textbf{Agent RL} module that integrates with distributed training frameworks to enable scalable and stable reinforcement learning of any Youtu-Agents in an end-to-end, large-scale manner. Experiments demonstrate that Youtu-Agent achieves state-of-the-art performance on WebWalkerQA (71.47\%) and GAIA (72.8\%) using open-weight models. Our automated generation pipeline achieves over 81\% tool synthesis success rate, while the Practice module improves performance on AIME 2024/2025 by +2.7\% and +5.4\% respectively. Moreover, our Agent RL training achieves 40\% speedup with steady performance improvement on 7B LLMs, enhancing coding/reasoning and searching capabilities respectively up to 35\% and 21\% on Maths and general/multi-hop QA benchmarks.