Late last Friday, Robby Stein and Rajan Patel from Digital Phablet announced on X that Google Search is now utilizing its newest AI model, Gemini 3.7 Flash. Although this model was first introduced on Thursday, Google didn’t specify its direct application within Search, describing it merely as their “most intelligent workhorse model yet for coding and agents.”
Initially, this feature is available in the AI Mode for Google AI Pro and Ultra subscribers in English, with a broader global rollout expected. Subscribers can access the updated model by clicking the plus icon within Search and selecting the new model.
Inquiries about potential integration into Search were made, revealing that as of now, Gemini 3.7 Flash isn’t officially built into Search’s interface. The following day, the team announced that the model is now being brought to Search to deliver more helpful responses. The rollout is happening worldwide and is accessible through AI Mode for eligible subscribers, simply by choosing the new model in the interface.
Robby Stein highlighted that the new model is better at following instructions and understanding user intent, leading to more accurate and helpful responses. Visual comparisons between the older model, Gemini 3.5 Flash-Lite, and the new Gemini 3.7 show that while local result accuracy remains similar, differences in top citation URLs are noticeable. Users have reported that the latest model can better understand complex queries, such as analyzing how Google reviews are updated or making more precise distinctions in search results.
Eventually, Gemini 3.7 Flash is expected to replace 3.5 Flash-Lite entirely in Search, especially as Google phases out the subscription-based access, making it available to all users in the future. Currently, the improvements are confined to paying subscribers in AI Mode, with no official announcement yet for broader deployment.