Google DeepMind Unveils Gemini 2.5 Flash for Precise Image Editing
Google DeepMind has integrated a new image editing model, Gemini 2.5 Flash Image Generation, into the Gemini app. This model, developed by Google, enables precise and localized edits via text input, significantly enhancing the app's image editing capabilities.
Gemini 2.5 Flash allows users to merge up to three images for complex compositions and stylistic transformations. It understands and visually represents simple causal relationships based on its world knowledge. The model can blur backgrounds, remove spots, add colors, or delete objects without manual selection tools, providing a user-friendly experience.
A key feature is character consistency, ensuring people, objects, or animals remain recognizable across different images. The model behaves similarly to GPT-4 in prompt execution, outperforming lower text-understanding image models. It's available within the Gemini app, using the 'Flash' language model instead of the previous 'Imagen' image model.
Gemini 2.5 Flash Image Generation is now available as a preview via the Gemini API, Google AI Studio, and Vertex AI, costing $30 per million output tokens, approximately $0.039 per image. This new feature radically alters images on request while keeping key elements recognizable, offering users a powerful tool for image editing and composition.
Read also:
- Mural at blast site in CDMX commemorates Alicia Matías, sacrificing life for granddaughter's safety
- Autonomous Vehicles Hit Manhattan & Brooklyn Streets, Sparking Controversy
- Increased energy demand counters Trump's pro-fossil fuel strategies, according to APG's infrastructure team.
- AI-Powered Transportation Stock's Possible Challenge to Tesla's Autonomous Dreams?
