This review article presents a comprehensive overview of multiscale computational and artificial intelligence (AI) models for both hard and soft composite materials. It discusses various modeling techniques, including molecular dynamics simulations, finite-element (FE) analyses, and machine learning (ML) and deep learning (DL) surrogate models. The article highlights the importance of multiscale modeling in understanding the complex behavior of composite materials, ranging from molecular to macroscopic scales. It covers the challenges associated with multiscale modeling, such as meshing, data variability, and material nonlinearity, and discusses the latest advancements in FE modeling, including meshless methods, hybrid ML and FE models, and nonlinear constitutive material models. The review also emphasizes the role of data-driven models in composite material research, which utilize advanced mathematical, numerical, and experimental data to develop accurate digital models. The article discusses the application of ML in composite modeling, including its benefits, challenges, and potential for future research. It also explores the evolution of multiscale modeling (MSM) and the different scales involved, from molecular to mesoscale and macroscale. The review highlights the importance of accurate material modeling in predicting composite behavior, including damage initiation, propagation, and failure. The article also discusses the use of hyperelastic and viscoelastic material models in composite modeling, as well as the emerging fractional viscoelastic model. The review concludes with a discussion on the future directions of multiscale modeling in composite materials, emphasizing the need for accurate and efficient models to advance the field of composite material research.This review article presents a comprehensive overview of multiscale computational and artificial intelligence (AI) models for both hard and soft composite materials. It discusses various modeling techniques, including molecular dynamics simulations, finite-element (FE) analyses, and machine learning (ML) and deep learning (DL) surrogate models. The article highlights the importance of multiscale modeling in understanding the complex behavior of composite materials, ranging from molecular to macroscopic scales. It covers the challenges associated with multiscale modeling, such as meshing, data variability, and material nonlinearity, and discusses the latest advancements in FE modeling, including meshless methods, hybrid ML and FE models, and nonlinear constitutive material models. The review also emphasizes the role of data-driven models in composite material research, which utilize advanced mathematical, numerical, and experimental data to develop accurate digital models. The article discusses the application of ML in composite modeling, including its benefits, challenges, and potential for future research. It also explores the evolution of multiscale modeling (MSM) and the different scales involved, from molecular to mesoscale and macroscale. The review highlights the importance of accurate material modeling in predicting composite behavior, including damage initiation, propagation, and failure. The article also discusses the use of hyperelastic and viscoelastic material models in composite modeling, as well as the emerging fractional viscoelastic model. The review concludes with a discussion on the future directions of multiscale modeling in composite materials, emphasizing the need for accurate and efficient models to advance the field of composite material research.