You can do a generalised mixed effects linear model with binomial outcome (ie a binomial test but with added random effects structure). But unless you want to introduce a richer random effects structure with more variables, it is overkill and overcomplicating things, and the result should be the same as t-tests.