Google DeepMind
Inventors of the Transformer, pioneering multimodal Gemini systems.
Gemini 3 Pro / Flash
OmniLatest Google flagship multimodal LLM with hyper-efficient Flash variants.
VaultGemma / Nested Learning
Privacy & Continual learningLaunch of the largest privacy-centric open model and the Nested Learning paradigm.
Gemini 2.5 Pro
ReasoningReasoning-focused Gemini model with long context and improved coding performance.
Imagen 4
ImageThe latest text-to-image series with unprecedented fidelity and prompt adherence.
Gemini 2.0 Flash
MultimodalExperimental release of the first Gemini 2.0 family model, optimized for speed and tool calling.
Veo
VideoGenerative video model for long-form content, later integrating native audio.
Gemini 1.5 Pro
Long ContextMassive context window breakthrough with near-perfect needle retrieval.
Gemini Pro / Ultra
MultimodalFirst major Google multimodal LLMs built from the start for diverse inputs.
BERT
EncoderBidirectional Encoder Representations from Transformers; foundational for NLP.
Transformer (Paper)
ArchitectureThe foundational architecture for all modern LLMs.