Google has introduced a new experimental tool called Mixboard, which allows users to create digital moodboards using AI. The service is based on the Gemini 2.5 Flash model and is available in open beta through Google Labs for users in the USA. Mixboard enables users to start a project with a text prompt or use ready-made templates, after which the AI selects images based on the given theme.
Users can add their own photos or generate new images through short descriptions, such as “show cups, bowls, and plates in Memphis style” or “plan an autumn party in my living room.” Boards can be edited, images combined, or altered using simple natural language commands. The system uses the Nano Banana model, noted for its ability to perform complex edits and create realistic images.
Mixboard is aimed at those looking for ideas for home decor, event planning, or creating creative collections for entertainment. The interface resembles familiar services like Pinterest or FigJam but offers more freedom by generating content from scratch. Additionally, users can click on quick options to refresh images or create similar designs.
Google reports that the AI can also generate text captions for images on the board. To start using Mixboard, users in the USA need to visit labs.google/mixboard. The company has created a separate community on Discord for sharing experiences and ideas.
