Google introduced Gemma 4, a compact, 'open' AI model sized to run on laptops and other personal devices, emphasizing on-device performance and accessibility. The smaller footprint is intended to bring generative capabilities to desktop apps and low-latency or offline scenarios, complementing larger Gemini-family models. Gemma 4 strengthens Google's multi-model approach by enabling local deployments that could reduce cloud costs and address privacy and responsiveness concerns raised by agentic and multimodal offerings.
Click a connection line between nodes to view confidence and evidence.