Google Updates Gemini Models
Curated by
elymc
3 min read
25,889
554
Google has unveiled significant updates to its Gemini AI models and Google Workspace offerings, expanding access to advanced AI capabilities while enhancing security features. As reported by TechRepublic, the standalone Gemini app is now included in Workspace Business, Enterprise, and Frontline plans, allowing millions more customers to leverage AI-powered tools with enterprise-grade data protections.
Gemini Model Enhancements
Google has made significant strides in enhancing its Gemini AI models, with the latest updates showing impressive performance gains. The company introduced a smaller variant of Gemini 1.5, called Gemini 1.5 Flash-8B, alongside improved versions of Gemini 1.5 Flash and Gemini 1.5 Pro
1
. These new models demonstrate increased capabilities across various benchmarks:
- Gemini 1.5 Flash has shown "huge gains" across the board
- Gemini 1.5 Pro exhibits notable improvements in math, coding, and handling complex prompts
- Both models feature 1 million token limits, allowing for processing of high-volume multimodal inputs
- The updated Gemini 1.5 Flash climbed from 23rd to 6th place in the Large Model Systems Organization's chatbot arena, matching Llama levels1
1 source
Gemini Access in Workspace
Google has expanded access to its Gemini AI capabilities across its Workspace suite, making the standalone Gemini app available as part of Workspace Business, Enterprise, and Frontline plans
1
. This integration allows millions of customers to leverage AI-powered tools with enterprise-grade data protections built-in1
. Key features of this expansion include:
- Enhanced administrative controls for managing Gemini's data retention practices1
- Assurance that Gemini will not use organizational data, user prompts, or generated responses to train or improve its models1
- Availability of Gemini in over 230 countries and territories2
- Option for businesses to unlock additional AI features with the Gemini for Workspace add-on, enabling direct integration with Gmail, Docs, Drive, and other applications3
1
.3 sources
New Security Features
A new AI-powered "Security Advisor" tool has been introduced for Workspace customers, delivering tailored security insights directly to IT administrators' inboxes
1
2
. This feature offers guidance on implementing threat defenses, account security measures, and data protection capabilities across Chrome, Gmail, and Google Drive. Additionally, an AI Security add-on, priced at $10 per user per month, enables automatic classification and protection of sensitive files in Google Drive using privacy-preserving AI models trained on organization-specific data3
4
.4 sources
Workspace App Innovations
Several innovative updates have been introduced to Google Workspace applications. A new AI-powered video creation app called Google Vids is set to launch in June, offering users an intuitive tool for work-related video content
1
. Meet's capabilities have been expanded with automatic caption translation in 52 additional languages, bringing the total to 692
. Sheets now features a tables function with pre-built templates, while Docs introduces tabs for better information organization and full-bleed cover images2
. Gmail mobile users can now benefit from voice input for the "Help me write" feature and a new "instant polish" function to refine drafts quickly2
.2 sources
Related
What are the key updates in Google Workspace for 2024
How does the new multisend mode in Gmail work
What are the advantages of using the new email layout feature
How can I use the Follow feature in Google Slides
What improvements were made to Google Takeout in the Admin Console
Keep Reading
Made by Google 2024 Recap
Google's Made by Google 2024 event showcased a range of innovative products and AI-powered features, headlined by the new Pixel 9 series and the introduction of Gemini Live, Google's advanced AI assistant. As reported by TechCrunch, the event emphasized Google's commitment to integrating cutting-edge AI capabilities across its hardware lineup, from smartphones to smart home devices, ushering in what the company calls "the Gemini era."
26,653
What Gemini AI Can Do: The Capabilities of Google’s Artificial Intelligence
Google's Gemini AI, a suite of powerful multimodal models, is designed to understand and generate various types of content, including text, code, audio, images, and video. As reported by Google DeepMind, Gemini's capabilities span complex tasks in math, physics, and coding, with the potential to transform how businesses operate and employees work across multiple industries.
5,277
GitHub Cuts AI Coding Deals
GitHub, owned by Microsoft, is enhancing its AI-powered coding assistant Copilot by integrating Google's Gemini and Anthropic's Claude 3.5 Sonnet models, offering developers increased flexibility and capabilities in code generation and chat functions, while also introducing new enterprise features and innovative tools like Project Spark for diverse user needs.
30,240