Google Launches Gemini 3.7 Flash at Half Price

Google has officially launched the AI model 'Gemini 3.7 Flash.' This new version strengthens support for coding assistance and AI agents—artificial intelligence systems that autonomously complete tasks on behalf of humans—with API usage fees set at half the standard rate through the end of 2026.

Notable is the fact that this update arrives just three weeks after the previous version, 'Gemini 3.6 Flash,' was released. Google attributes this short update cycle to developer feedback and algorithmic improvements. As competition in AI model development intensifies, the Flash series is increasingly positioned as an 'everyday use model' emphasizing practical utility in business settings, with frequent enhancements continuing.

On the pricing front, during the period through the end of 2026, rates are set at $0.75 per million input tokens and $3.75 per million output tokens. Starting January 1, 2027, these will increase to $1.50 and $7.50 respectively. This discount period can be viewed as a grace period for enterprise teams deploying large-scale code processing or business automation agents to validate actual cost-saving benefits.

In terms of performance, Google describes Gemini 3.7 Flash as 'the most capable practical model for coding and agents to date.' Specifically, the company highlights improved handling of obstacles encountered during task execution, enhanced ability to accurately comprehend the intent behind instructions, and strengthened capacity to execute multi-step plans more carefully. Google DeepMind has also announced improvements in debugging and problem-solving accuracy, as well as generation of more practical web layouts.

While the previous version 3.6 Flash was optimized to 'reduce reasoning steps and unnecessary tool calls,' 3.7 Flash shifts emphasis toward 'allocating sufficient thought to planning and raising execution quality.' In scenarios where AI agents are used for business automation, executing each step correctly proves more important than arbitrarily reducing steps—a strategic shift with significant practical implications.

Meanwhile, 'Gemini 3.5 Pro,' the higher-tier model, had no release date included in this announcement. According to Reuters, the model has previously been described as being in the testing phase with partner companies, but no specific public launch timeframe has been provided. Axios also notes that 3.7 Flash has been released ahead of the long-anticipated Pro model.

Google's decision to continue updating the Flash series without waiting for the flagship Pro model's announcement represents a strategy to maintain market presence as a practical tool for developers and enterprises. The half-price positioning facilitates cost comparison with competing models, making the key question whether Google can maintain equivalent performance and cost competitiveness at standard pricing after the discount period ends in 2027.