Report: Google's Gemini 3.5 Pro launch delayed

Alphabet has postponed the launch of Gemini 3.5 Pro by several months. The main reason is that the model's coding capabilities fell short of internal expectations. When this news broke, Alphabet's stock fell 4% in a single day.

Recently released new AI models from OpenAI and Meta are evaluated as superior to Google's current offerings in code generation capability. OpenAI's GPT-5.6 Sol delivers 54% higher token efficiency in agentic coding tasks, and Meta's Muse Spark 1.1 was described by the company's AI chief as "the most powerful model for agentic and coding tasks."

Code generation has now established itself as a key use case for AI model providers such as Anthropic, OpenAI, and Z.ai. In a statement, an Alphabet spokesperson said the company is developing "a range of models that are cost-efficient and quick to deploy," and that 3.5 Pro is currently in the testing phase with partners.