GLM-4-Flash demonstrates exceptional performance in evaluations across datasets involving semantics, mathematics, reasoning, coding, and knowledge. We offer a variety of large models, each with different features and price points. Image generation model is billed based on requests, and other models are billed based on tokens.