Google Unveils Gemini 3 Flash as Standard AI Model Worldwide
On December 17, 2025, Google officially released Gemini 3 Flash, setting it as the standard AI model for both the Gemini app and Google Search’s AI Mode globally. This move signifies a notable escalation in Google’s rivalry with OpenAI.
Gemini 3 Flash is touted as providing “frontier intelligence built for speed at a fraction of the cost.” It delivers Pro-level reasoning capabilities with a marked increase in efficiency. The model achieved a score of 90.4% on the GPQA Diamond benchmark and 33.7% on Humanity’s Last Exam. This performance rivals that of larger frontier models such as Gemini 3 Pro and OpenAI’s GPT-5.2, while operating at a speed three times faster than Gemini 2.5 Pro.
The new model shines in multimodal tasks, allowing users to analyze videos, images, audio, and text with a heightened level of understanding. Since the launch of Gemini 3, Google reports processing over 1 trillion tokens per day on its API. Major enterprise customers such as Salesforce, Workday, Figma, JetBrains, Cursor, and Harvey have already begun utilizing the model.
For developers, Gemini 3 Flash can be accessed through Google AI Studio, Vertex AI, and the newly introduced Google Antigravity platform. The pricing is set at $0.50 per million input tokens and $3.00 per million output tokens. This release comes hot on the heels of OpenAI’s launch of GPT-5.2, further heating up the ongoing AI race between the two tech giants.
Source: TechCrunch