According to reports, Google's artificial intelligence system processed more than 980 trillion tokens in June, more than double the number in May. According to Logan Kilpatrick, a Google product manager, and Demis Hassabis, CEO of DeepMind, tokens are short text segments used by AI models to understand or generate responses.
This significant increase not only reflects the rise in Google AI usage, but more importantly, it demonstrates the efficient application of reasoning models like Gemini Flash 2.5. Gemini Flash 2.5 can process more tokens, thereby providing more accurate responses.
According to a report by Artificial Analysis, the number of tokens used by Flash 2.5 has increased by about 17 times compared to its previous version. Although the cost of performing reasoning tasks has also increased by 150 times, its ability to handle large amounts of information is undoubtedly the key to this data surge.