Introducing FrontierCode

Today’s coding benchmarks have established that models can write correct code, but the question we should really be asking is: can models actually write good code?