California, USA, August 14, 2026 — Google has launched Gemini 3.7 Flash, just three weeks after introducing Gemini 3.6 Flash, with the company positioning the new model as its most capable “workhorse” model yet for coding and AI agents.
Google said Gemini 3.7 Flash is the result of developer feedback and algorithmic improvements that are expected to contribute to future models. Compared with Gemini 3.6 Flash, the new version brings upgrades across software engineering, knowledge work and web development.
For software engineering, Google says Gemini 3.7 Flash is better equipped to handle tasks such as debugging and resolving issues. In web development, the model can produce more functional layouts and feature-complete applications with fewer prompts.
The model also offers improved reasoning and accuracy for knowledge-intensive areas including finance, law and biosciences, according to Google.
Gemini Spark, Google’s 24/7 AI agent launched earlier this year, is also switching to Gemini 3.7 Flash from today. Google says the upgrade will make Spark more efficient for knowledge work while improving accuracy and output quality when handling complex workflows involving multiple skills.
