Pollination is a critical ecosystem function for sustaining biodiversity. However, pollinators and the services they provide are threatened by landscape-altering anthropogenic activities across the globe. Habitat loss and fragmentation, introduction of invasive species, chemical use, and urbanization have been shown to impact pollination. Pollinator foraging behavior is thought to be largely a function of available floral rewards, therefore, understanding the role of resource distributions in pollinator abundance and behavior within disturbed landscapes is a key piece of information for conservation. Fine scale information on floral resource distribution across disturbed landscapes is lacking in most systems. Here we demonstrate how existing presence-only species distribution modeling techniques (i.e., Maximum entropy modeling [MaxEnt]) can be combined with widely available environmental information to create resource landscapes for both pollinator communities and specific pollinators of interest. This model is the first of its kind, making possible simultaneous visualization of fine-scale resource configuration and quantity across abroad spatial extent. We tested this method to build caloric landscapes using tropical hummingbird-plant system in Costa Rica. We found that our MaxEnt models performed well on independent data for all 13 flower species we examined. Our landscape-scale caloric map showed that available calories within each 35m² pixel ranged from 0 to greater than 30,000 across our study region. Our model provides the possibility of predicting pollinator movement and abundance based upon resource supply. As its parameters are flexible, it is broad in its potential applications. The flexibility in calibration to desired resource landscapes permits applying the model to other pollinator-plant systems. We hope that this model will complement the current ecologist’s toolbox, aiding in ensuring the continuation and health of pollinator systems.