The hypothesis that large brains allow animals to produce novel behaviour patterns is supported by the correlation between brain size, corrected for body size, and the frequency of foraging innovations reported in the literature for both birds and primates. In birds, foraging innovations have been observed in over 800 species, and include behaviours that range from eating a novel food to using tools. Previous comparative studies have quantified innovativeness by summing all reports of innovative behaviour, regardless of the nature of the innovation. Here, we use the variety of foraging innovations recorded for birds to see which of two classic hypotheses best accounts for the relationship between innovativeness and brain size: the technical intelligence hypothesis or the opportunistic-generalism intelligence hypothesis. We classified 2182 innovation cases into 12 categories to quantify the diversity of innovations performed by each of 76 avian families. We found that families with larger brains had a greater repertoire of innovations, and that innovation diversity was a stronger predictor of residual brain size than was total number of innovations. Furthermore, the diversity of technical innovations displayed by bird families was a much better predictor of residual brain size than was the number of food type innovations, providing support for the technical intelligence hypothesis. Our results suggest that the cognitive capacity required to perform a wide variety of novel foraging techniques underpins the positive relationship between innovativeness and brain size in birds. We include a summary of innovation data for 803 species as Supplementary Material.