Google’s new Gemini 3.7 Flash AI model boosts issue resolution and debugging for developer platforms and agent frameworks.
Google launched Gemini 3.7 Flash on Thursday, targeting software coding and automated business workflows. However, the company remained silent on the release date for its flagship Pro model.
Investors are closely monitoring the timeline for Google’s next flagship Pro model, viewing it as a critical benchmark of whether Google DeepMind can keep pace with rivals like Anthropic and OpenAI.
Google is positioning the model as a cost-effective choice for businesses building autonomous AI systems that can plan tasks, navigate software, and execute multi-step workflows with minimal human supervision.
Launched just three weeks after Gemini 3.6 Flash, version 3.7 delivers improved performance in software coding specifically in debugging, issue resolution, and generating production-ready code, according to a Google blog post.
To spur adoption, Google is offering Gemini 3.7 Flash at a discounted introductory rate of $0.75 per million input tokens and $3.75 per million output tokens through the end of the year half the original cost of Gemini 3.6 Flash.
The model is rolling out immediately on Gemini Spark, Google’s subscription AI agent service available to Google AI Pro and Ultra subscribers in over 160 countries.
Google co-founder Sergey Brin has recently urged top AI staff to focus entirely on developing the company’s Gemini model, as parent company Alphabet works to close the gap with competitors, according to an exclusive report from Reuters.
See also: Claude Users Push Back as Anthropic’s New Text Watermarks Threaten Workplace and Classroom AI Use
Last week, Google announced a major leadership overhaul at its DeepMind AI division. Chief Executive Demis Hassabis stepped aside to become chair, handing operational control to his deputy, Koray Kavukcuoglu. Meanwhile, Gemini’s original technical co-leads left the company to launch a new startup.
CEO Sundar Pichai mounted a robust defense of Google’s AI strategy during July’s earnings call, pushing back on concerns that delays to its flagship model meant the company was falling behind rivals in coding and agentic execution.
