Gemini 3.7 Flash の導入価格引き下げを発表

2 Min Read

Google launched Gemini 3.7 Flash, just three weeks after releasing Gemini 3.6 Flash, a testament to how fast the AI race is moving and to the application that vendors see as most compelling to enterprises.

Google said 3.7 Flash, released on Thursday, is more effective than its predecessor in coding tasks such as debugging and IT problem resolution. The model also has higher code accuracy and generates production-ready code, Google said. Other capabilities include improved reasoning and accuracy for applications such as finance, law and bioscience.

Google is offering the model at an introductory price of $0.75 per million input tokens and $3.75 per million output tokens, lower than the prices for models released by Anthropic, OpenAI and SpaceXAI in recent weeks.

The introductory price is Google’s response to the price war in the AI market. Vendors are introducing new models weekly, trying to outprice one another. The pricing battle stems from enterprises' concerns about the high cost of using AI tokens and models for their AI goals and workflows.

Related:IBM, OpenAI Partner to Accelerate Enterprise AI

The Pricing War

For Google, though, pricing has long been a key point of competitiveness, said Arun Chandrasekaran, an analyst at Gartner.

“One area where Google is trying to really impress upon the customer is: ‘It's not just the performance; it's price, and we have a very, very competitive price performance related to other frontier labs,’” Chandrasekaran said.

As Google tries to compete against OpenAI and Anthropic, pricing is a way to entice some enterprise developers and many digital-native startups away from competitors, he added.

The Focus on Coding

Beyond pricing, Google’s emphasis on coding with Gemini 3.7 Flash shows it recognizes this as a successful use case that OpenAI and Anthropic have pursued as well.

“Google has been a little slow in creating very highly capable coding models,” Chandrasekaran said. “Coding agents are one of the fastest growing use cases with AI, so in some ways Google is trying to get to where it wants to be.”

Not only has coding been successful for frontier labs, but open model providers are also targeting the coding use case. For instance, Meta’s Muse Glimmer is focused on helping users handle multi-step coding. DeepSeek-V4-Pro is built for reasoning-heavy coding tasks. Moonshot AI’s Kimi K3 is optimized for long-horizon code generation.

“Some of this is also an acknowledgment that the open weight models are becoming a legitimate contender in the coding category and overall reasoning as well,” Chandrasekaran said. “That's also another thing Google has to contend with.”

Related:Prompt: Wall Street Is Coming for AI Infrastructure

He added that, with Google less assertive in the AI market in 2026, the rapid release of Gemini 3.7 Flash and introductory pricing is the tech giant’s way of trying to reclaim mindshare among enterprise developers.

About the Author

News Writer, AI Business

Esther Shittu has covered AI technologies and industry trends since 2021. As co-host of the Targeting AI podcast, she talks with experts, thought leaders and practitioners exploring critical AI developments. Before AI Business, she wrote for SearchEnterpriseAI, the New York Daily News, Bklyner and the Brooklyn Daily Eagle. When she's not diving deep into the world of AI, she spends her time on passion projects and raising her three daughters.