About our Google Gemini news
Latest news on Google Gemini, Google's upcoming artificial intelligence system, which is being developed by the Google DeepMind division.
Google Gemini, launched on 6 December 2023, marks a significant step forward in AI capabilities and applications. Gemini is designed to be Google's largest and most capable AI model. It's highly flexible, running efficiently on diverse platforms from data centers to mobile devices.
Gemini 1.0 has been optimized in three different versions, catering to various needs:
1. Gemini Ultra: Suitable for highly complex tasks, being the largest and most capable model.
2. Gemini Pro: Best suited for scaling across a wide range of tasks.
3. Gemini Nano: Most efficient for on-device tasks, ideal for Android users and app development.
Gemini Ultra surpassed human experts in Massive Multitask Language Understanding (MMLU), scoring 90.0% in academic benchmarks. It outperformed existing models in 30 of the 32 widely-used academic benchmarks for large language models (LLMs).
Gemini's design is natively multimodal, pre-trained on different modalities like text, images, audio, etc., and then fine-tuned with additional data. This helps it seamlessly understand and reason about varied inputs.
Gemini Pro is being integrated into various Google products, including Bard, for advanced reasoning and understanding. Gemini Nano is being brought to Pixel smartphones, enhancing features like the Summarize function in the Recorder app and Smart Reply in Gboard.
From December 13, 2023, developers and enterprise customers can access Gemini Pro via the Gemini API in Google AI Studio or Google Cloud Vertex AI. Android developers will be able to use Gemini Nano via AICore in Android 14 on Pixel 8 Pro devices.
Extensive safety checks are ongoing for Gemini Ultra. It will be made available for select customers and developers for early experimentation, with plans for a broader rollout to developers and enterprise customers in the upcoming year.