Google has unveiled Gemini 2.5 Flash Image, a new AI image model that allows users to edit visuals using natural language, aiming to compete with OpenAI's ChatGPT. This tool provides functions like merging and editing photos with text prompts while maintaining visual accuracy and character consistency. It enables tasks such as pose changes and multi-image fusions. The model, which costs $30 per million output tokens, can seamlessly integrate multiple images and leverage world knowledge for improved results. Google aims to integrate advanced image editing into Gemini's offerings, expanding the tool's access via platforms like OpenRouter and fal.ai, thereby broadening its reach to developers worldwide. In contrast, OpenAI's ChatGPT, which already boasts over 700 million weekly users, remains a dominant player in the field as Google seeks to close the gap with its 400 million monthly active users. The company reinforces its commitment to responsible AI by naming all outputs with an invisible watermark to clarify that they are AI-generated.

Source 🔗