Alphabet Inc. shares rose 0.5% heading into Wednesday’s close, down from gains of more than 2%, after a Bloomberg report described internal employee skepticism about Gemini 4, the company’s upcoming artificial intelligence model. The report said Gemini 4 performed well on industry benchmarks but was less effective on practical tasks, including some coding work. Google rejected the characterization, saying it would be inaccurate to describe the model as underperforming in coding and pointing to comments from Koray Kavukcuoglu, head of Google DeepMind, who said last week that he was encouraged by its performance. Google had also planned, then abandoned, a June release of Gemini 3.5 Pro. Some employees told Bloomberg that Anthropic’s Fable and OpenAI’s Astra were advancing faster, while others believed Gemini 4 had caught up with leading AI labs. A Google employee familiar with development said there was "large consensus" internally that Gemini 4 was at the frontier and denied problems with real-world coding tasks. Gemini models support products including Search, Maps, Gmail and Chrome, each of which has more than a billion users.