Gemini launches the "Deep Think" mode, enhancing reasoning and code generation capabilities.

According to news from HashiChain, Gemini has officially launched the "Gemini 3 Deep Think" mode, which enhances its reasoning and code generation capabilities. The new mode utilizes iterative reasoning loops, allowing for more detailed and complex code output, suitable for applications in visualization, prototyping, and experimental scenarios. Currently, this feature has been made available to Google AI Ultra subscription users, who can select the "Thinking" mode and enable the "Deep Think" mode for use.