A flood of flagship releases includes Gemma 4, DeepSeek V4, and Kimi K2.6. These updates push the boundaries of open-weight performance across multiple architectures. Researchers are now applying the CAISI V4 assessment to benchmark these systems. This surge provides practitioners with high-tier alternatives to closed-source models for complex deployment tasks.