Google's AI lab and maker of the Gemini multimodal model family.
Category: Foundation Models · Type: Foundation model provider · Website: https://deepmind.google
Overview
Google DeepMind builds the Gemini models, natively multimodal systems with very long context windows. Gemini is delivered to developers through Google AI Studio and Vertex AI, and underpins agent features across Google's ecosystem. The mid-2026 lineup includes Gemini 3.1 Pro and Gemini 3.5 Flash.
Key Features
- Gemini natively multimodal models (text, image, audio, video)
- Industry-leading long context windows
- Delivered via Google AI Studio and Vertex AI
- Deep integration with Google Workspace and Cloud
Role in the Agentic AI Stack
A frontier 'brain' especially strong on multimodal and long-context tasks; the model layer beneath Google's enterprise agent platform Vertex AI.
Part of Agentic AI Landscape → Foundation Models