Google's shares fell after the launch of its latest artificial intelligence (AI) model, "Gemini 3.5 Pro," was delayed by several months from schedule.
According to Bloomberg on the 16th, Google has failed to release Gemini 3.5 Pro for more than two months since unveiling it at its annual developer conference "Google I/O 2026" in May this year. Google missed its launch schedule originally planned for last month and remains far from ready internally. As a result, the launch is reported to be pushed back by several months.
Bloomberg said the launch delay is due to problems with the model's coding performance. The report is interpreted to mean that Gemini 3.5 Pro is still struggling to gain a competitive edge, as rivals including OpenAI compete to release models that significantly improve capabilities for work, particularly coding for developers. Google reportedly updated its model training data at the end of last month to improve coding performance but failed to achieve proper results.
There are also suggestions that the fundamental cause of these launch setbacks lies in Google's complex organizational structure. Unlike competitors that focus mainly on AI model development alone, Google inevitably suffers from lower efficiency because interests of its vast product lineup, including search, cloud, maps and YouTube, are intertwined.
According to AI performance analysis firm Artificial Analysis, Gemini 3.1 Flash ranks 13th in the global rankings of major models. It ranks lower than competing models such as Claude Fable5, GPT-5.6 Sol, as well as Grok4.5 and Muse Spark1.1, and even below its own Gemini 3.1 Flash (12th).
On news of the launch delay, Google's shares fell 4.44% in regular trading that day to $354.46.
Google said, "We are rapidly releasing a wide range of models while maintaining high cost efficiency for our customers," adding, "We are currently testing 3.5 Pro and improved Flash models with our partners," Bloomberg reported.
