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

Google launched Gemini 3.8 Flash only weeks after Gemini 3.7 Flash, saying the new model performs more reasoning steps and can call tools iteratively on complex tasks. While introductory token prices remain unchanged, Google warns the model may use more tokens, especially at higher effort levels, increasing total costs for some developers.

The release reflects a broader shift toward models that trade efficiency for stronger reasoning and agent-like tool use. For developers, headline pricing is becoming less informative as actual AI costs increasingly depend on token consumption, task complexity, and model behavior.