Mistral: Mistral Large 4

Mistral Large 4, a frontier multimodal model from Mistral AI, demonstrates competitive response times, ranking in the 54th percentile across four benchmarks. Its pricing is also competitive, placing it in the 46th percentile. Notably, the model exhibits exceptional reliability with a 100% success rate across all benchmarks, indicating consistent and usable responses. In terms of performance across categories, Mistral Large 4 excels significantly in Mathematics, achieving 97.0% accuracy and ranking in the 96th percentile, showcasing its strong analytical capabilities. It also performs well in Coding and Reasoning, with accuracies of 91.8% (62nd percentile) and 90.0% (63rd percentile) respectively. While its Instruction Following accuracy is moderate at 62.0% (51st percentile), it still provides a solid foundation for agentic workloads. The model's key strengths lie in its high accuracy in complex mathematical and coding tasks, coupled with its outstanding reliability. Its primary area for potential improvement appears to be in the nuanced precision required for advanced instruction following.