Google releases the Gemini 1.0 family
Google introduces the natively multimodal Gemini 1.0 family—Ultra, Pro and Nano—with Pro and Nano entering products immediately and developer APIs following a week later.
What happened
Event details
Google released the first Gemini model family in three sizes: Ultra for complex tasks, Pro for broad deployment and Nano for on-device use. Gemini was pretrained jointly across text, images, audio and video rather than assembled from separately trained single-modality components. Google did not disclose parameter counts for the three sizes. Availability was staged by model. Gemini Pro entered the English version of Bard on launch day, and Gemini Nano powered features on Pixel 8 Pro. The Gemini Pro API opened through Google AI Studio and Vertex AI on December 13. Gemini Ultra was still undergoing safety evaluation and did not reach users through Gemini Advanced until February 2024.
Assessment
Why it matters
Gemini 1.0 unified Google's cloud, product and on-device model strategy around native multimodal training, then reached a broad developer and consumer ecosystem.
Availability
Access notes
Pro and Nano entered products at launch; the Pro API opened on December 13, while Ultra was not yet broadly available.