Google launches Gemini 3.7 Flash with significant coding improvements

Google unveiled Gemini 3.7 Flash, a new iteration of its core AI model arriving just weeks after version 3.6 Flash. The update incorporates engineering refinements and developer feedback to strengthen coding capabilities and agentic performance. The company is introducing competitive introductory pricing to challenge rival offerings. Performance gains span multiple evaluation benchmarks. The model demonstrated substantial improvements in coding tasks, with FrontierCode 1.1 Main jumping from 34.4 to 43.6 percent and DeepSWE v1.1 rising from 49 to 65.3 percent. Web development performance also improved, with WebDev Arena scores climbing from 1,538 to 1,588. Document processing capabilities showed progress as well, with the GDP.pdf benchmark increasing from 22 to 34 percent. Business workflow automation tests revealed similar advances, with AutomationBench reaching 30.4 percent compared to 3.6 Flash's 17 percent performance.