What happened?

Google unveiled its new flagship AI model Gemini 4 with impressive benchmark scores. However, a report claimed the model "struggles" in some real-world use cases.

Google's response

Google said it stands by Gemini 4's performance. The company expressed confidence in the model's capabilities, responding directly to the criticism.

Why it matters

A performance debate starting right after a new flagship model's launch raises the question of which model users should trust. The gap between benchmark scores and everyday experience is a frequently discussed topic in the AI world.