Solvent selection plays a critical role in pharmaceutical crystallization, affecting product quality, manufacturability, regulatory compliance, and environmental impacts. With the growing emphasis on sustainability and life cycle assessment (LCA) in regulatory frameworks, solvent design must go beyond thermodynamic considerations to actively incorporate environmental responsibility. However, current practices still heavily rely on expert heuristics and trial-and-error experimentation, which are time-consuming and difficult to scale. Although methods such as Computer-Aided Molecular Design and other commercial tools are available, they often experience high complexity, limited usability, and inadequate integration of sustainability metrics. To address these limitations, we developed a hybrid, data-driven framework with two primary goals: (a) to identify green solvents or binary mixtures that meet solubility, sustainability, and uncertainty criteria for a given API, and (b) to provide a user-friendly, computationally efficient platform for pharmaceutical engineers. The framework includes a solubility database of over 60,000 data points across 1,183 solvent systems, integrated with sustainability assessments based on ReCiPe 2016 and the GSK Solvent Sustainability Guide. Machine learning models—including a multi-task polynomial regression network, a temperature-adjusted solubility predictor, and a binary solvent model—were combined with Monte Carlo uncertainty quantification to enable a robust solvent recommendation. The entire workflow is encapsulated in SolECOs, an interactive GUI that supports model execution, solubility curve visualization, and dynamic sustainability analysis. A case study on cytarabine demonstrated the platform’s value under different cooling profiles and sustainability priorities. The model consistently identified 1,2-dimethylbenzene as a primary solvent, with polar co-solvents like dichloromethane and ethanol enhancing solubility. However, high predicted solubility did not always align with favorable environmental scores, underscoring the need to balance performance and sustainability. Case studies on four APIs, including cytarabine, under varying cooling gradients and sustainability priorities validated the model’s accuracy and robustness, further demonstrating SolECOs’s ability to support reliable, goal-oriented, and sustainability-aware solvent selection.