Abstract:
Three weeks after the release of the previous model, Google began launching Gemini 3.8 Flash on September 2, local time. This is also the third update to the Gemini Flash series in three months. Google said that Gemini 3.8 Flash is its most intelligent Flash model currently, designed for long-cycle software engineering, automated agents and complex enterprise workflows.
Judging from the information released so far, Gemini 3.8 Flash has outperformed the previous generation model in multiple benchmark tests. However, Google has not announced specific test results in the relevant introduction.
Similar to the previous generation model, the knowledge deadline for Gemini 3.8 Flash is March 2026 in some areas; but in other areas, users may find that the knowledge held by the model is still only updated until January 2025.
Google has once again launched a limited-time discount price: until December 31, 2026, the price of input tokens is US$0.75 per 1 million coins, and the price of output tokens is US$3.75 per 1 million coins.
Currently, Gemini 3.8 Flash is available to Google AI subscribers in the Gemini application, and is also available in Google Antigravity and AI Studio.
Comments