Overview

• Developers can access the preview now through the Gemini API, Google AI Studio and Vertex AI, with a stable release expected in the coming weeks.
• The model maintains character and style consistency across multiple generations to keep subjects identifiable in varied scenes.
• Conversational, prompt-driven editing enables precise modifications to existing images through natural language.
• Multi-Image Fusion can combine features from several input images into a single, coherent output.
• Google enforces automated content filtering and SynthID digital watermarking to provide provenance and mitigate misuse.