Klear-AgentForge: Forging Agentic Intelligence through Posttraining Scaling

TL;DR

This paper introduces Klear-Qwen3-AgentForge, an open-source pipeline for training high-performance agentic language models capable of tool use and coding, achieving state-of-the-art results among similar-sized models.

Contribution

It presents a comprehensive, open-source training pipeline combining supervised fine-tuning and reinforcement learning for agentic models starting from Qwen3-8B.

Findings

Achieves state-of-the-art performance among similar-sized models.

Remains competitive with larger models.

Effective for diverse agentic tasks.

Abstract

Despite the proliferation of powerful agentic models, the lack of critical post-training details hinders the development of strong counterparts in the open-source community. In this study, we present a comprehensive and fully open-source pipeline for training a high-performance agentic model for interacting with external tools and environments, named Klear-Qwen3-AgentForge, starting from the Qwen3-8B base model. We design effective supervised fine-tuning (SFT) with synthetic data followed by multi-turn reinforcement learning (RL) to unlock the potential for multiple diverse agentic tasks. We perform exclusive experiments on various agentic benchmarks in both tool use and coding domains. Klear-Qwen3-AgentForge-8B achieves state-of-the-art performance among LLMs of similar size and remains competitive with significantly larger models.