The Year in AI Papers
The year/Independent research

Paper 2512.04987

Nex-N1: Agentic Models Trained via a Unified Ecosystem for Large-Scale Environment Construction

Nex-AGI Team, :, Yuxuan Cai, Lu Chen, Qiaoling Chen, Yuyang Ding, Liwen Fan, Wenjie Fu, Yufei Gao, Honglin Guo, Pinxue Guo, Zhenhua Han, et al.

Published
Dec 2025
Research lab
Independent
Citations
15
GitHub
Not linked

01 In brief

Summary

The paper introduces Nex-N1, a series of agentic models trained using a unified ecosystem (NexAU, NexA4A, NexGAP) for large-scale environment construction.

NexAU is a modular runtime for scalable agent frameworks, NexA4A automatically generates diverse agent hierarchies from natural language, and NexGAP generates end-to-end agentic trajectories using real MCP tools.

The models are trained on diverse and complex interactive environments.

Results show Nex-N1 outperforms open-source models of comparable size and achieves competitive performance against proprietary models on benchmarks like SWE-bench, GAIA 2, and BFCL.

The models also demonstrate robustness across different agent frameworks and excel in practical tasks like web development and deep research.

The authors open-source the ecosystem and model weights.

02 From the paper

Abstract

The evolution of Large Language Models (LLMs) from passive responders to autonomous agents necessitates a fundamental shift in learning paradigms -- from static imitation to incentive-driven decision making. However, this transition is significantly impeded by the lack of scalable infrastructure capable of constructing high-quality interaction signals for effective policy learning. To address this, we introduce a comprehensive method designed to systematically scale the diversity and complexity of interactive environments. Our method realizes this scaling by addressing three orthogonal dimensions: (1) Complexity: NexAU, a flexible agent framework that supports building complex agent hierarchies via simple configurations; (2) Diversity: NexA4A automatically generates diverse agent hierarchies from natural language to cover infinite domains; and (3) Fidelity: NexGAP bridges the simulation-reality gap by integrating dynamic real-world environment for grounded trajectories synthesis. We train Nex-N1 upon the diverse and complex interactive environments established by our infrastructure. Empirical results on benchmarks such as SWE-bench and tau2 demonstrate that Nex-N1 consistently outperforms SOTA open-source models and achieves competitive performance against frontier proprietary models on complex agentic tasks. We open-source the Nex ecosystem and model weights to facilitate further research.