Frontier multimodal model release puts video understanding and live voice in focus

Summary

The model shows stronger cross-modal reasoning, pushing real-time assistants, education, and creator tools into the next phase.