--- title: "Google accelerates the AI race: introduces Gemini 3.7 Flash with a 50% price cut" description: "🚀 Google introduced Gemini 3.7 Flash just 3 weeks after the release of version 3.6. The new model is twice as cheap ($3.75 per 1M tokens) and significantly smarter: performance growth in tests by 18.8%. This is a response to the failures of the Pro series and an attempt to regain leadership in the AI agent race. 💻📉 #Google #Gemini #AI" date: 2026-08-15T20:19:07.000Z lang: en url: https://xab.info/en/posts/google-accelerates-ai-race-gemini-3-7-flash-price-cut tags: [google, gemini, artificial-intelligence, ai-agents, tech-news, machine-learning] publisher: "XAB.info" --- # Google accelerates the AI race: introduces Gemini 3.7 Flash with a 50% price cut ![Logo and name of the new Google AI model Gemini 3.7 Flash on a blue background](https://xab.info/media/2026/08/16/google-gemini-3-7-flash-udeshivlenie-i-rost-moshchnosti/google-gemini-3-7-flash-udeshivlenie-i-rost-moshchnosti-1.webp) ## 🎯 Key Points - Google released Gemini 3.7 Flash 3 weeks after version 3.6, accelerating the update cycle. - The model cost has been halved: $3.75 per 1 million output tokens compared to $7.50 for the predecessor. - Performance increased: 20th place in Agent Arena and +18.8% in DeepSWE tests. - The model is positioned as the best tool for programming and AI agents. In the face of fierce competition in the artificial intelligence market, Google is taking decisive steps to strengthen the position of its Gemini model lineup. On August 16, 2026, the company unveiled a new version — Gemini 3.7 Flash. This release comes against the backdrop of the corporation's efforts to rectify the situation following the controversial launch of previous versions, including Gemini 3.5 Pro. The new model is positioned as a key tool for software code development and the creation of autonomous AI agents. ### Acceleration strategy: from 3.6 to 3.7 in three weeks The speed of updates has become one of the main characteristics of the current stage of Google AI product development. Gemini 3.7 Flash was released just three weeks after the launch of version 3.6 Flash. Such a compressed interval between updates, which would previously have been unthinkable for major technology companies, signals a paradigm shift: Google is moving towards a model of continuous and rapid evolution of its algorithms. Experts link this to the need to regain lost ground and compensate for delays in launching flagship models of the Pro series. ### Economic breakthrough: cost reduction with increased power One of the most significant aspects of the new release was a radical restructuring of pricing. Until the end of 2026, the cost of using Gemini 3.7 Flash will be $0.75 per 1 million input tokens and $3.75 per 1 million output tokens. For comparison, the previous version, Gemini 3.6 Flash, cost $7.50 per million output tokens. Thus, Google has not only increased performance but also halved the product's cost, making it more accessible to developers and corporate clients. ### Technical achievements and leadership in benchmarks The new model demonstrates impressive results in independent tests. In the Agent Arena ranking, which evaluates AI capabilities in performing complex tasks, Gemini 3.7 Flash took a confident 20th place, whereas its predecessor (3.6 Flash) was only in 35th position. In the DeepSWE test, which checks software problem-solving skills, the novelty scored 65.5%. This is 18.8% higher than the result of the previous version. Google states that Gemini 3.7 Flash has become their most intelligent working model, specifically tuned for production code generation. ### Ecosystem integration and new features In addition to improving the model's 'brain', Google has expanded its functional capabilities. According to sources, Gemini 3.7 Flash has received deep integration with Google Workspace, allowing users to work with documents, spreadsheets, and calendars directly within the chat interface. A watermark toggle feature was also added, giving developers more control over AI-generated content. These updates are aimed at making the model not just a text generation tool, but a full-fledged assistant in workflows. ## 🔍 Fact-Check Verification - [Google releases Gemini 3.7 Flash with Workspace integration and watermark toggle](https://www.msn.com/en-us/news/other/google-releases-gemini-37-flash-with-workspace-integration-and-watermark-toggle/gm-GMD48C3069) - Подтверждает интеграцию с Workspace и дату релиза. - [Google Brings Gemini 3.7 Flash To AI Mode In Search](https://www.searchenginejournal.com/google-brings-gemini-3-7-flash-to-ai-mode-in-search/585923/) - Подтверждено по источнику searchenginejournal.com - [Google Launches Gemini 3.7 Flash With Huge Upgrades & Lower Costs](https://www.techjuice.pk/google-launches-gemini-3-7-flash-with-huge-upgrades-lower-costs/) - Подтверждает снижение стоимости и улучшение характеристик. - [Google introduces Gemini 3.7 Flash with production-code focus](https://www.msn.com/en-us/news/technology/google-introduces-gemini-3-7-flash-with-production-code-focus/ar-AA2a7ZZV) - Подтверждает интеграцию с Workspace и дату релиза. ## ❓ FAQ ### Q: When was the Gemini 3.7 Flash model released? **A:** The model was introduced on August 16, 2026, just three weeks after the release of version 3.6 Flash. ### Q: How much does it cost to use Gemini 3.7 Flash? **A:** The cost is $0.75 per 1 million input tokens and $3.75 per 1 million output tokens until the end of 2026. ### Q: What is the difference between Gemini 3.7 Flash and 3.6 Flash? **A:** The new model is twice as cheap, faster, and smarter: it took 20th place in Agent Arena (compared to 35th for the predecessor) and showed an 18.8% increase in the DeepSWE test.