This paper examines price competition on iFood using a novel high frequency dataset built through large-scale web scraping. I estimate a structural logit demand model for three representative markets and find high price elasticities and low markups. The estimates show temporal stability while reflecting meaningful heterogeneity across markets and external conditions such as rainfall. Counterfactual simulations demonstrate that coordinated price increases fail to raise aggregate profits, indicating strong competitive pressure among sellers. Overall, the results provide quantitative evidence on how delivery platforms shape pricing incentives and constrain market power.