Google DeepMind releases Gemma 4: Frontier multimodal intelligence on device
Action Required
Organizations need to migrate to Gemma 4 to leverage its advanced multimodal capabilities and improved performance, requiring code changes and potentially infrastructure adjustments.
AI Impact Summary
Google DeepMind is releasing Gemma 4, a new family of multimodal models designed for on-device intelligence. These models offer significant improvements over previous versions, including support for image, text, and audio inputs, and are available with Apache 2 licenses. The models come in four sizes, with impressive performance benchmarks, and leverage architectural innovations like Per-Layer Embeddings and a Shared KV Cache for efficiency. This release represents a major step towards truly open and accessible frontier AI models.
Affected Systems
- Date
- Date not specified
- Change type
- capability
- Severity
- high