PANews reported on April 10 that according to TechCrunch, Google released a new AI model, Gemini 2.5 Flash, which is designed for efficient processing, high throughput and low-cost scenarios. The model will soon be launched on the Google Vertex AI platform, supporting users to adjust the balance between speed, accuracy and cost, and is suitable for real-time tasks such as customer service and document parsing. As an "inference-based" model, Gemini 2.5 Flash can perform self-verification before answering. Google also plans to deploy the model to a local environment in Q3 this year, and implement compliance through Google Distributed Cloud and Nvidia Blackwell systems.