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

Core Insights - Google has launched the Gemini 3 Flash model, aiming to compete with OpenAI's offerings, and it will be the default model in the Gemini app and AI mode in search [1][4] Model Performance - The Gemini 3 Flash model shows significant improvements over its predecessor, Gemini 2.5 Flash, and matches the performance of other leading models like Gemini 3 Pro and GPT 5.2 in certain benchmarks [2] - It scored 33.7% on Humanity's Last Exam benchmark, compared to 37.5% for Gemini 3 Pro, 11% for Gemini 2.5 Flash, and 34.5% for GPT-5.2 [3] - On the MMMU-Pro benchmark, the new model achieved an 81.2% score, outperforming all competitors [3] Consumer Rollout - The Gemini 3 Flash model is now the default in the Gemini app globally, replacing the previous version, while users can still opt for the Pro model for specific tasks [4] Features and Capabilities - The new model excels in identifying multimodal content and can provide answers based on various inputs, such as videos, sketches, and audio recordings [5] - It has improved understanding of user intent and can generate more visual responses, including images and tables [5] Enterprise and Developer Availability - Companies like JetBrains, Figma, Cursor, Harvey, and Latitude are already utilizing the Gemini 3 Flash model, which is accessible through Vertex AI and Gemini Enterprise [10] - The model is available for developers in a preview mode via API and in Antigravity, a new coding tool [10] Pricing and Efficiency - The pricing for the new model is set at $0.50 per 1 million input tokens and $3.00 per 1 million output tokens, which is higher than the previous model but offers better performance and speed [13] - The Gemini 3 Flash model is three times faster than Gemini 2.5 Pro and uses 30% fewer tokens on average for thinking tasks, potentially leading to cost savings [13][15] Competitive Landscape - The release of the Gemini 3 Flash model comes amid increased competition with OpenAI, which has also launched new models in response to Google's market share growth [16] - Google emphasizes that the ongoing development of models across the industry is pushing the boundaries of AI capabilities [17][18]