Gemini Glossary

Google DeepMind's proprietary model family, used across the Gemini app, Google products, and Vertex AI.

Gemini is the proprietary model family produced by Google DeepMind. Google announced the first generation in December 2023 in three sizes: Ultra, Pro, and Nano. Later releases expanded the family across long-context, multimodal, reasoning, and agentic uses; Gemini 1.5 introduced a context window of up to one million tokens, and the Gemini 3 series was current in 2026.

For this Dictionary, Gemini matters at three levels: as a peer of GPT and Claude in the closed-model market; as the proprietary sibling of the open-weights Gemma line; and as the model family Google integrates into consumer products and its Vertex AI platform. The name therefore refers both to models and to the consumer application built around them. Keeping those layers separate prevents product features from being mistaken for properties of one model.

Sources

• Google, “Introducing Gemini”, December 2023.
• Google, “Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context”, February 2024.
• Google, “A new era of intelligence with Gemini 3”, 2026.

See also

• Google DeepMind
• Gemma — the open-weights sibling
• GPT, Claude — peer frontier model families
• Demis Hassabis — the CEO of Google DeepMind

Return to Dictionary All Entries (A–Z) For Students Other Writing Capstone 2.0