Reasoning Models

AI models specifically designed for complex logical and mathematical reasoning, such as OpenAI's o1/o3 series and DeepSeek R1. These models allocate additional compute at inference time to think through problems step-by-step, achieving significantly higher accuracy on challenging benchmarks.