These models actually outperform traditional methods on many fronts, including accuracy a lot of the time. They are technically generative AI models, but they're definitely not LLMs.