Google launches Gemini 3 Flash, makes it the default model in the Gemini app

Google has unveiled the Gemini 3 Flash, a faster and more cost-effective version of its Gemini 3 AI model released last month. This new model is now the default choice in the Gemini app and AI mode on Google Search, aiming to challenge OpenAI’s dominance in the AI space.

Performance and Benchmarks

Building on the advancements of Gemini 2.5 Flash released six months prior, the Gemini 3 Flash significantly surpasses its predecessor in benchmarks. It competes closely with leading AI models like Gemini 3 Pro and OpenAI's GPT 5.2 in various tests. For example, on the Humanity’s Last Exam benchmark, which evaluates expertise across domains, Gemini 3 Flash scored 33.7%, compared to Gemini 3 Pro's 37.5% and GPT-5.2's 34.5%. It also outperformed all competitors in the multimodality and reasoning benchmark, MMMU-Pro, with an 81.2% score.

Global Consumer Rollout

Google has replaced Gemini 2.5 Flash with Gemini 3 Flash as the default model globally in the Gemini app. While users can still select the Pro model for specialized tasks like math and coding, Gemini 3 Flash excels at understanding multimodal inputs such as images, audio, and videos. For instance, users can submit short clips, sketches, or audio recordings for analysis or interactive responses like quizzes.

The model also demonstrates better understanding of user intent and can produce more visual and structured responses, including the use of images and tables. Additionally, users can create app prototypes within the Gemini app using prompts powered by Gemini 3 Flash.

Enterprise and Developer Access

Major companies including JetBrains, Figma, Cursor, Harvey, and Latitude are already integrating Gemini 3 Flash. It is accessible via Vertex AI and Gemini Enterprise platforms. For developers, the model is available in a preview API and through Google's new coding tool, Antigravity.

Gemini 3 Pro, a higher-tier model, boasts a 78% score on the SWE-bench coding test, ranking just behind GPT-5.2. The Flash model's speed and efficiency make it ideal for video analysis, data extraction, and visual question answering, supporting rapid and repetitive workflows.

Model Pricing and Efficiency

Pricing for Gemini 3 Flash is $0.50 per million input tokens and $3.00 per million output tokens, a slight increase from Gemini 2.5 Flash's rates. However, Google highlights that Gemini 3 Flash outperforms the Gemini 2.5 Pro model while being three times faster and consuming 30% fewer tokens on average during cognitive tasks, which may result in overall savings for certain uses.

Tulsee Doshi, Google's Senior Director and Head of Product for Gemini Models, described Gemini 3 Flash as a "workhorse model" designed for efficiency and widespread, bulk task handling.

Industry Competition

Since the Gemini 3 launch, Google processes over 1 trillion tokens daily through its API. The release of Gemini 3 Flash coincides with heightened competition from OpenAI, which issued an internal “Code Red” memo following a decline in ChatGPT traffic and responded by releasing GPT-5.2 and a new image generation model.

While Google did not directly comment on this rivalry, it noted that the ongoing innovation from all companies is pushing AI technology forward and encouraging new standards and benchmarks across the industry.

For more details, visit the original TechCrunch article: Google launches Gemini 3 Flash, makes it the default model in the Gemini app.