Art&Graft evolves Gemini from technical showcase to living system

Art&Graft has worked in close partnership with Google DeepMind to shape the visual language for the launch of Gemini 3. The London-based design and motion studio, which had previously collaborated with the team on the identity for Gemini 2.0, was brought back to evolve that system for a product whose audience had broadened significantly. Gemini sits across a huge range of Google’s products, and the visual language needed to shift accordingly – retaining technical sophistication while becoming approachable enough for a much wider spectrum of users. The result of the collaboration is a particle system delivered across hundreds of assets, alongside motion toolkits, templates and style guides designed to work from large presentation environments down to mobile screens. AI visual language has a well-documented tendency toward certain tropes. Art&Graft and the Google DeepMind team were conscious of this territory from the outset. “Neural networks, abstract data streams and similar motifs have become quite established, and if you’re not careful they bring a lot of baggage with them,” Founder & ECD Mike Moloney notes. For a brand that sits at the centre of so many different products and services, landing somewhere distinctive felt essential. Together, the teams arrived at the idea of framing Gemini as a kind of thoughtful co-creator that’s ambient, intelligent and always present. From that positioning, everything else followed. The visual system suggests an intelligent presence that is constantly evolving and ready to respond, rather than visualising computation directly. Senior Motion Designer Gulya Voronova describes the underlying metaphor as rooted in nature. The project uses the image of a shoal of fish – dot and pill elements moving together, changing direction, responding to their surroundings while maintaining clear structure. This creates a visual language that feels alive and intuitive, making complex technology easier to grasp.