Google's Gemini 3.7 Flash Boosts AI Model Capabilities

Google has released Gemini 3.7 Flash, an upgraded version of its AI model designed to handle complex tasks and workflows. One of the biggest upgrades is the integration with Google's cloud-based agent, Spark.

The company claims that Spark will benefit from the new model's improved tool use and multi-step planning capabilities, which should result in fewer failed tasks and less manual intervention for users.

Gemini 3.7 Flash has already shown impressive results on benchmarks, scoring 30.4% on AutomationBench, compared to 17% for its predecessor Gemini 3.6 Flash. It also outperformed GPT-5.6 Terra at 23.6% and Claude Sonnet 5 at 10.7%.

The model's pricing is another notable aspect, with introductory rates of $0.75 per million input tokens and $3.75 per million output tokens until December 31. This makes it significantly cheaper than its competitors, including Claude Sonnet 5 and GPT-5.6 Terra.