Google's Gemini 3.5 Pro is reportedly months behind schedule as coding results fall short

Google’s flagship Gemini 3.5 Pro is running months behind schedule as the company works to improve the model’s capabilities, particularly in coding, according to people familiar with the matter.

The slip carries an immediate market cost. Alphabet shares fell nearly 4% on July 16, erasing roughly $200 billion in market capitalization after the delay was reported.

Gemini 3.5 Pro is positioned as Google’s most powerful AI model, and it had been expected in June. Chief Executive Sundar Pichai referenced it at the company’s I/O developer conference in May. Google updated the training data feeding Gemini last month to sharpen its coding performance, but the results fell short of internal goals, the people said.

The stakes are competitive as much as technical. The delay has frustrated some Google engineers, researchers and managers who worry the company is losing ground to Anthropic and OpenAI, whose coding-focused models have set the pace this year. Coding has become the clearest commercial proving ground for frontier models, and a flagship that misses internal targets there is a costly signal to enterprise buyers weighing which model to standardize on.

Google has not publicly confirmed a delay or commented on its internal targets, and the June timetable was never announced as a formal ship date. The account rests on people describing private deliberations, so the shortfall is measured against goals only Google can see.

The test now is whether Gemini 3.5 Pro ships this quarter with coding gains Google can demonstrate, or slips further while rivals iterate.