Gemini 3.7 Flash lands with coding gains and undercuts its three-week-old predecessor's price by 50%
Google released Gemini 3.7 Flash, claiming superior coding and agent performance compared to Claude Sonnet 5 and GPT-5.6 Terra while reducing costs by 50% versus its predecessor.
Google has launched Gemini 3.7 Flash, following the release of 3.6 Flash by only three weeks. The company positions this iteration as its most capable model for coding tasks and AI agent workflows.
According to Google's internal benchmarks, the new model outperforms competitors like Claude Sonnet 5 and GPT-5.6 Terra. Simultaneously, the pricing structure has been adjusted to undercut the previous version by half, aiming for better cost-efficiency.
This rapid iteration cycle suggests an aggressive strategy to dominate the developer and enterprise sectors. By lowering inference costs while claiming performance gains, Google targets widespread adoption for agentic applications and software development tools.
The focus on coding and agents aligns with current industry trends where efficiency and specialized capabilities drive model selection. Competitors will likely need to respond to these benchmark claims and pricing adjustments in the coming weeks.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.