On August 13 local time, Google announced the launch of the Gemini 3.7 Flash model, designed specifically for programming and agent scenarios.
The release comes only about three weeks after the launch of Gemini 3.6 Flash. Google said that this version was released directly in response to developer feedback and a series of algorithmic innovations. Gemini 3.7 Flash delivers major improvements in software engineering, knowledge work, and web development workflows.
In addition, through the end of this year (until December 31, 2026), Gemini 3.7 Flash will be available at an introductory promotional price (Gemini 3.6 Flash will receive the same discount), just half the original per-million-token price of 3.6 Flash:
Input: $0.75 per million tokens (approximately RMB 5.1 at the current exchange rate)
Output: $3.75 per million tokens (approximately RMB 25.3 at the current exchange rate)
Starting today, Gemini Spark, available to Google AI Pro and Ultra subscribers in more than 160 countries and regions, will officially integrate Gemini 3.7 Flash. The model upgrade significantly improves Spark's efficiency in handling knowledge work and optimizes its ability to use tools in the Google Workspace suite, delivering greater accuracy and higher-quality output for complex, cross-skill workflows.
In addition, Gemini 3.7 Flash introduces new safety defenses that guard against misuse in the chemical, biological, radiological, and nuclear (CBRN) fields as well as cyberattacks, while ensuring that legitimate and beneficial applications can proceed smoothly.