Google has launched Gemini, a generative AI model developed by Google DeepMind.

The company claims that the system outperforms OpenAI's GPT-4 at 30 out of 32 model benchmarks.

Google Cloud
– Sebastian Moss

Gemini comes in three variants - Nano, a smaller version that can be run natively on Android devices; Pro, which will power a number of Google AI services and is the backbone of the Bard platform; and Ultra, the most powerful large language model developed by Google.

Ultra is set to launch next year.

Developers and enterprise customers will be able to access Gemini Pro through Google Generative AI Studio or Vertex AI over Google Cloud starting on 13 December.

While Nano and Pro will be initially language only, Ultra will also be able to handle images, video, and video.

"We’ve made great progress on Gemini so far and we’re working hard to further extend its capabilities for future versions, including advances in planning and memory, and increasing the context window for processing even more information to give better responses," Demis Hassabis, CEO and co-founder of Google DeepMind, said in a blog post.

More in Cloud & Hyperscale