Overview

• Gemini 2.5 Flash is available in preview through the Gemini app, API, Google AI Studio, and Vertex AI, offering developers early access to its features.
• The model introduces hybrid reasoning, allowing developers to toggle its 'thinking' function on or off for tailored performance.
• A 'thinking budget' feature lets developers allocate computational resources, optimizing cost and latency for complex tasks.
• Priced at $0.15 per million input tokens and $0.60/$3.50 per million output tokens (without/with reasoning), it aims to deliver cost-efficient reasoning capabilities.
• Google is gathering developer feedback during the preview phase to refine its dynamic reasoning features for broader deployment.