Google has introduced Gemini 2.5 Flash Image, its latest state-of-the-art model designed to advance AI-driven image creation and editing. The upgrade delivers greater creative control, higher-quality visuals, and seamless integration with real-world knowledge, enabling developers and enterprises to build more powerful visual applications.
The model introduces several new capabilities, including multi-image fusion, allowing users to blend multiple photos into a single composition; character consistency, ensuring subjects retain their identity across edits and scenes; and prompt-based editing, where simple natural language commands can remove elements, adjust poses, or modify environments. Leveraging Gemini’s deep world knowledge, the model also supports use cases in education, product design, real estate, and more.
Gemini 2.5 Flash Image is now available through the Gemini API, Google AI Studio, and Vertex AI for enterprise customers. Priced at $30 per million output tokens (approximately $0.039 per image), it maintains the cost-efficiency of earlier versions while offering enhanced creative precision.
Also Read: Microsoft’s AI Chief Warns of ‘Seemingly Conscious AI’ as a Dangerous Turning Point
To make adoption easier, Google has updated AI Studio’s build mode, enabling developers to quickly prototype, remix, and deploy applications. Partnerships with OpenRouter.ai and fal.ai will expand global access, bringing the model to millions of developers.
In line with responsible AI practices, every image generated with Gemini 2.5 Flash Image will include an invisible SynthID digital watermark to identify AI-created or edited content.