Google Chooses Gemini 3 Flash as Default Model: What You Need to Know!

Google is stepping up its game. Just one month after unveiling Gemini 3 Pro, the tech giant from Mountain View is launching Gemini 3 Flash, an optimized version designed for speed without sacrificing performance. The company is not just announcing this new model; it’s also deploying it immediately as the default in the Gemini app, replacing Gemini 2.5 Flash.

The release is part of a fierce competition with OpenAI. Just a few days ago, in a state of high alert, OpenAI launched GPT-5.2 against Gemini 3. Google is now responding by not only incorporating Gemini 3 Flash in its consumer app but also in Google Search through AI Mode (not available in France), in its developer tools like Google AI Studio, Antigravity, and Android Studio, as well as in Vertex AI for businesses.

Three Times Faster Than Its Predecessor

Gemini 3 Flash positions itself as the ideal working model for developers and businesses. According to Google, it outperforms Gemini 2.5 Pro in performance while being three times faster. This speed does not prevent the model from excelling in complex tasks that require advanced reasoning and multimodal understanding, as per Google’s promise.

The model’s efficiency is also reflected in its token consumption. For thought-intensive tasks, Gemini 3 Flash uses on average 30% fewer tokens than Gemini 2.5 Pro, notes the Mountain View firm. This represents a significant economic advantage for businesses that handle large volumes of queries.

However, in terms of pricing, the model costs 0.50 dollars per million input tokens and 3 dollars per million output tokens. This pricing is slightly higher than Gemini 2.5 Flash (0.30 and 2.50 dollars respectively), but justified by noticeably improved performance, Google assesses.

Gemini 3 Flash Becomes the Default Model in the App

Effective immediately, Gemini 3 Flash is replacing Gemini 2.5 Flash as the default model in the Gemini app, which is freely accessible to all users worldwide. Subscribers can still select Gemini 3 Pro through the model selector for more complex tasks, such as in mathematics or coding.

In terms of features, the model leverages its multimodal capabilities to analyze videos, images, and audio files. For instance, it can analyze a short video to provide improvement tips, identify an ongoing drawing, or generate a personalized quiz from an audio recording. Google also highlights the ability to create functional applications directly via voice command, without any programming knowledge. Additionally, Opal, the firm’s vibe coding tool, is also integrated into the Gemini app on the web.

Available for Developers and Integrated into Google Search

For developers, Gemini 3 Flash is available immediately via the Gemini API in Google AI Studio, the agent development platform Antigravity launched last month, as well as in Gemini CLI and Android Studio. Businesses can access it through Vertex AI and Gemini Enterprise.

The model is also being deployed in Google Search, where it becomes the default model for AI Mode worldwide. However, this conversational search feature is still not available in France. In the United States, users also have access to Gemini 3 Pro directly in Search, as well as the image generation model Nano Banana Pro.