Gemini 3.7 Flash: a practical look at “workhorse” coding and agent behavior
Gemini 3.7 Flash (Aug 13, 2026) is positioned as Google’s most intelligent “workhorse” model for coding and agents, with an introductory $0.75 per 1M input tokens and $3.75 per 1M output tokens through the end of 2026. The focus is better execution: fewer retries, stronger instruction fidelity, and improved tool-using behavior for multi-step workflows. ([blog.google](https://blog.google/innovation