Based on the figure, you should be looking at models with strain softening behavior rather than elasto-plastic models. I would prefer to use models with fracture energy concept included to capture the effect of localization.
Thank you for these answers and i have the pleasure of sharing my opinion on modeling the behavior shown in the attached figure.
So, if it is a compression test of a composite material, the model of Drucker Prager seems to me able to reproduce the observed behavior.
Drucker-Prager model is available in most finite element codes used in civil engineering field, especially those devoted more specifically to geotechnical applications. Indeed,this model is a simple first idealization of the behaviour of frictional-cohesive materials, like composite and other granular materials.
Finally, I renew my thanks and invite you to give some examples of other models available in finite element codes that you considered efficient for such behavior.
For uniaxial compression with strain softening and considering the fracture energy concept, the best reference would be the paper of Prof. Hikaru Nakamura from Nagoya University entitled "Compressive Fracture Energy and Fracture Zone Length of Concrete." For high strength concrete, the work of Prof. Akiyama entitled "Stress-Averaged Stain Model for High Strength Concrete" would be a great reference.