Data & Training
Data, RL, and Evaluation
Generate training data, integrate with RL frameworks, and evaluate agent performance across 30+ benchmarks
Data Generation
Chain-of-thought synthesis, self-instruct pipelines, and tool-integrated reasoning data.
Learn moreRL & Environments
Reinforcement learning integration with AReaL, Verl, and Slime frameworks.
Learn moreBenchmarking
30+ integrated benchmarks from SWE-Bench to GAIA for comprehensive evaluation.
Learn more