Google has just unveiled Gemini, its most advanced model boasting sophisticated multimodal reasoning capabilities. Gemini is engineered for versatility, being finely optimized for three distinct sizes: Gemini Ultra, Pro, and Nano, making it compatible with a wide range of platforms, from data centers to mobile devices.
The exciting news is that Gemini is now making its way to Bard in what is Bard’s most significant upgrade to date. The integration of Gemini into Bard will take place in two phases. Starting today, Bard will incorporate a specially tailored version of Gemini Pro in English, enabling more advanced reasoning, enhanced planning, and improved comprehension. Additionally, early next year, we will introduce Bard Advanced, offering you exclusive access to our most cutting-edge models and capabilities, beginning with Gemini Ultra.
Gemini Pro In Google BARD
You can start experiencing Gemini Pro in Bard right away. Prior to its public release, we rigorously subjected Gemini Pro to a battery of industry-standard benchmarks. Impressively, Gemini Pro outperformed GPT-3.5 in six out of eight benchmarks, including the crucial MMLU (Massive Multitask Language Understanding) and GSM8K, which evaluates grade school math reasoning.
Furthermore, Gemini Pro in Bard has been finely tuned to excel in areas such as understanding, summarization, reasoning, coding, and planning. The results are remarkable, as confirmed by blind evaluations with third-party raters. Bard now stands out as the preferred free chatbot when compared to leading alternatives, a testament to its enhanced capabilities.
In an exciting collaboration, we teamed up with YouTuber and educator Mark Rober to put Bard with Gemini Pro to the ultimate test: crafting the most accurate paper airplane. Witness how Bard elevated the creative process to new heights in the video above.
Starting today, you can try out Bard with Gemini Pro for text-based prompts, with support for other modalities coming soon. Initially available in English across more than 170 countries and territories, it will expand to include additional languages and regions, including Europe, in the near future.
Gemini Ultra in an advanced version of Bard early next year
Looking ahead, be prepared for Gemini Ultra, our most expansive and capable model, engineered for tackling highly complex tasks and rapidly processing diverse forms of information, including text, images, audio, video, and code.
Your opportunity to experience Gemini Ultra will come through Bard Advanced, a cutting-edge AI experience within Bard that grants access to our premier models and capabilities. We are diligently conducting comprehensive safety checks and will soon launch a trusted tester program before opening up Bard Advanced to a broader audience early next year.
This development aligns perfectly with our steadfast commitment to responsible AI. We’ve integrated safety features into Bard based on our AI Principles, including the introduction of contextual assistance, such as Bard’s “Google it” button, which makes it easier to double-check its responses. As we continue to fine-tune Bard, your feedback will play a crucial role in our ongoing improvements.
With Gemini, we are one step closer to realizing our vision of making Bard the ultimate AI collaborator in the world. We are thrilled to keep incorporating the latest advancements into Bard and eagerly anticipate witnessing how you leverage it to create, learn, and explore. Don’t hesitate to try Bard with Gemini Pro today.