The latest model can process video, audio, and images simultaneously in real-time conversations. Early benchmarks show 40% improvement on complex reasoning tasks. Available to Plus subscribers immediately.
Read original article →