Google has launched Gemini 3.7 Flash on Thursday, August 13, its second low-cost coding model in three weeks, while the much-anticipated Gemini 3.5 Pro promised to developers since May remains delayed. The company is pitching the new model as its strongest workhorse for coding and agents, but the rollout may frustrate enterprise users expecting the flagship model that was supposed to compete more directly with Anthropic and OpenAI.
The main selling point is price. Gemini 3.7 Flash will cost $0.75 per million input tokens and $3.75 per million output tokens through the end of the year. That is half the launch price of Gemini 3.6 Flash. However, OpenAI’s Luna version of the GPT-5.6 lineup is priced even lower at $0.20 per million input tokens and $1.20 per million output tokens, underscoring the intensifying AI price war.
Google backed the launch with benchmark improvements over 3.6 Flash. On the DeepSWE v1.1 coding test, performance rose from 49.0% to 65.3%, while FrontierCode 1.1 Main increased from 34.4% to 43.6%. The model’s WebDev Arena score climbed from 1,538 to 1,588. Knowledge work also improved: the GDP.pdf benchmark score rose from 22.0% to 34.0%, and AutomationBench moved from 17.0% to 30.4%, according to Senior Director Tulsee Doshi.
Access is limited. Developers can use Gemini 3.7 Flash via the Gemini API, AI Studio, Android Studio, Google Antigravity, and the Gemini Enterprise platform. Individual access is restricted to Google AI Pro and Ultra subscribers. The standard Gemini chatbot continues to use 3.6 Flash, while the upgrade to 3.7 has made it easier for Spark to handle multi-step tasks across Gmail, Google Docs, and other Workspace apps. In the broader AI race, Gemini 3.7 Flash still trails leading models on the Artificial Analysis Intelligence Index, where Anthropic’s Claude Opus 5 currently leads.
