Google says its new Gemini 3.8 Flash model ‘works harder’ but might cost more

Google launched a new AI model called Gemini 3.8 Flash that performs more reasoning steps and uses tools iteratively compared to its predecessor. While the per-token pricing remains the same as the previous version, the model may consume more tokens to maximize performance, potentially resulting in higher overall costs for users. Early assessments indicate the model shows significant improvements for software engineering and autonomous AI agents, outperforming competitors on several benchmarks. Google also released a specialized version called Gemini 3.8 Flash Cyber through a new program limited to governments and trusted partners.