Snake venoms are primarily used to subjugate prey, and consequently, their evolution has been shown to be predominantly driven by diet-related selection pressure. Venoms tend to be more lethal to prey than non-prey species (except in cases of toxin resistance), prey-specific toxins have been identified, and preliminary work has demonstrated an association between the diversity of diet classes and that of toxicological activities of whole venom. However, venoms are complex mixtures of many toxins, and it remains unclear how toxin diversity is driven by diet. Prey-specific toxins do not encompass the molecular diversity of venoms, and whole venom effects could be driven by one, few, or all components, so the link between diet and venom diversity remains minimally understood. Here, we collated a database of venom composition and diet records and used a combination of phylogenetic comparative methods and two quantitative diversity indices to investigate whether and how diet diversity relates to the toxin diversity of snake venoms. We reveal that venom diversity is negatively related to diet diversity using Shannon’s index but positively related using Simpson’s index. Since Shannon’s index predominantly considers the number of prey/toxins, whereas Simpson’s index more strongly reflects evenness, we provide insights into how the diet–venom diversity link is driven. Specifically, species with low diet diversity tend to have venoms dominated by a few abundant (possibly specialised) toxin families, whereas species with diverse diets tend to ‘hedge their bets’ by having venoms with a more even composition of different toxin classes.