Introducing Gemini 3.7 Flash
deepmind ·
Google has launched Gemini 3.7 Flash, positioning it as its most capable foundational model for software development and autonomous agents. Arriving just three weeks after its predecessor, the update incorporates direct developer feedback alongside new algorithmic refinements to boost performance across coding, web design, and specialized professional domains. The model demonstrates marked advancements in generating production-ready code, executing complex document analysis, and automating business workflows, according to recent benchmark evaluations. Developers will also benefit from a smoother workflow characterized by more reliable multi-step planning and reduced need for manual corrections. To encourage rapid adoption, Google is offering an introductory rate that halves the token costs compared to the previous version through the end of the year. Additionally, the architecture will soon power Gemini Spark for premium subscribers, enhancing its daily productivity capabilities within Google Workspace applications. Alongside these performance upgrades, the release includes strengthened security protocols designed to prevent misuse in sensitive technical fields. The model is now accessible through various developer tools, enterprise platforms, and consumer applications worldwide.
deepmind