Google Unveils Gemini Enterprise AI Agents

Welcome to another edition of AI PlanetX.

Google launches Gemini Enterprise for custom AI agents; Microsoft unveils Nvidia GB300 supercluster for OpenAI; Reflection AI raises $2B to build an open alternative to closed labs.

• Hottest AI News

• Top AI & SaaS Tools

• AI Tutorial: Run OpenAI’s GPT-OSS Models Locally

• Top AI & Tech News

• AI Art Spotlight

• Prompt of the Day: Make Next 90 Days Legendary

• AI Video Tutorial

• Course of the Day: AI Coding Masterclass

Hottest AI News

DeepMind

Google Launches Gemini Enterprise to Help Businesses Build Custom AI Agents

Google is taking a major step into enterprise AI with fresh Gemini subscriptions that let businesses create custom AI agents without coding. This move puts Google in direct competition with OpenAI and Microsoft as the race to dominate corporate AI heats up.

Details:

• Google has launched two Gemini subscription tiers: Gemini Enterprise at 30 per user monthly for large companies, and Gemini Business at 21 for smaller teams. Both plans let employees build AI agents that connect with Box, Microsoft, and Salesforce, while offering ready-made Google agents for coding, data, and customer engagement

• These tiers build on Agentspace, Google’s agent-building tool released in December. Existing users will get upgrades during their current contracts. A built-in security layer called Model Armor scans and blocks unsafe AI chats

• The rollout comes just days after OpenAI added third-party apps to ChatGPT and Amazon revealed Quick Suite, a similar no-code agent platform. Google Cloud CEO Thomas Kurian says industries from consulting to manufacturing are testing Gemini, with Virgin Voyages as an early adopter

Google’s approach aims to make AI usable for businesses of all sizes, providing a secure, intuitive platform that could accelerate enterprise adoption.

Typing is a thing of the past

Typeless turns your raw, unfiltered voice into beautifully polished writing - in real time.

It works like magic, feels like cheating, and allows your thoughts to flow more freely than ever before.

With Typeless, you become more creative. More inspired. And more in-tune with your own ideas.

Your voice is your strength. Typeless turns it into a superpower.

Download for Mac today

Microsoft

Microsoft Azure Unveils World's First Nvidia GB300 Supercomputer Cluster for OpenAI

Microsoft Azure has unveiled the world’s first Nvidia GB300 supercomputing cluster — a historic leap in AI infrastructure. Built in partnership with Nvidia, it delivers unmatched computing power tailored for OpenAI’s most advanced workloads.

Details:

• The NDv6 GB300 VM series marks a huge leap in supercomputing, powered by 4,600+ Nvidia Blackwell Ultra GPUs connected via Quantum-X800 InfiniBand. Each liquid-cooled system combines 72 GPUs and 36 Nvidia Grace CPUs into one highly efficient unit

• It delivers extreme performance — 37 terabytes of memory and 1.44 exaflops of FP4 Tensor Core power per VM. The setup is optimized for agentic AI systems, with major upgrades in networking and memory for large-scale performance

• The deployment dramatically transforms AI development timelines, reducing OpenAI's model training duration from weeks to just days, while highlighting the strategic partnership between Microsoft and Nvidia

This launch represents more than just powerful hardware - it demonstrates Microsoft Azure and Nvidia's shared commitment to optimizing every component of modern AI data centers.

Top AI & SaaS Tools

• Fox Signals (Life-time Deal): AI trading signals app that scans crypto, stocks, and forex to detect patterns and deliver high-accuracy, profit-focused alerts

• Dia: AI browser for writing, learning, planning, shopping, and translation — with inline edits, tutoring, summaries, and task help (macOS) [F-R-E-E]

• Huxe App: Creates audio briefings from your emails, calendar events, and news so you can listen on the go [F-R-E-E]

• Qwen-Image-Edit-2509: Major rebuild of the image editor for creators, designers, and AI tinkerers — precise single- and multi-image editing with built-in ControlNet [F-R-E-E]

• ElevenLabs UI: MIT-licensed UI library with 22 customizable components and examples for AI audio/voice agents [F-R-E-E]

The Gold standard for AI news

AI will eliminate 300 million jobs in the next 5 years.

Yours doesn't have to be one of them.

Here's how to future-proof your career:

• Join the Superhuman AI newsletter - read by 1M+ professionals

• Learn AI skills in 3 mins a day

• Become the AI expert on your team

Start learning AI now

AI Tutorial

Run OpenAI’s GPT-OSS Models Locally Offline on Your Computer

Want ChatGPT-style AI without monthly fees, internet dependency, or privacy concerns? This tutorial will show you how to run OpenAI’s powerful open-source models directly on your computer, completely at no cost.

1. Understand your options

   You can run the models in two ways:

   • Ollama: a terminal-based command-line interface.

   • LM Studio: a graphical app with a ChatGPT-like experience.

2. Check system requirements

   • Storage: at least 13 GB available.

   • Memory: 16 GB of RAM or more recommended.

3. Choose your method

   Decide whether you want the simplicity of a GUI (LM Studio) or the flexibility of a terminal (Ollama).

4. Install Ollama

   Download Ollama from its official site and complete the installation. Once installed, open your terminal to begin.

5. Run model

   In your terminal, enter:

   ollama run gpt-oss:20b

   This will download the 13 GB model. Afterward, you can run it completely offline.

6. Install LM Studio (If you don't want to use Ollama)

   Download and install LM Studio as you would any desktop application.

7. Set up LM Studio

   • Launch the app.

   • Go to the Discover section.

   • Find and download the GPT-OSS-20B model.

8. Start chatting

   Once the model is ready, you can chat in a familiar interface that feels like ChatGPT.

9. Use Playground as alternative

   If local installation isn’t practical for your computer, visit the GPT-OSS Playground online. It lets you test the models directly in your browser without heavy downloads.

Note: For smoother performance, close background apps before running the model. If you’re using LM Studio, enable hardware acceleration in the settings for quicker responses.

Top AI & Tech News

• Reflection AI founded by two ex-DeepMind researchers, raised $2B at an $8B valuation to build an open-frontier AI lab as an alternative to closed labs like OpenAI and Anthropic

• Rishi Sunak has taken paid senior advisory roles with US tech companies Microsoft and Anthropic, alongside other paid positions at Goldman Sachs and as a paid speaker

• Together AI introduced ATLAS, an adaptive speculator system that boosts real-time LLM inference by up to 400% vs. current baselines

• Adobe introduced latest AI agents for B2B marketing teams, including Audience, Journey, and Data Insights

AI Art Spotlight

Model: Gemini 2.5 Flash Image (Nano Banana)

Prompt:

Based on my uploaded photo create a Ultra-realistic 8K close-up portrait of a rugged man with wet skin glistening under dramatic cinematic lighting. Intense focus on one eye with sharp blue iris detail, wet strands of hair falling across the face, moisture droplets illuminated by warm orange and golden highlights. Strong texture of skin and beard captured with extreme clarity, high contrast between light and shadow, evoking a powerful and dramatic atmosphere. Hyper-detailed, photorealistic, cinematic style

Prompt of the Day

Prompt to Make the Next 90 Days Legendary

This prompt helps you pick one clear, high‑impact goal and create a simple 90‑day plan to get real results. It strips away busywork and excuses, forces you to face the hard truth about what’s holding you back, and makes you commit to one focused action you can do every day.

Top AI Video Tutorial

How to Connect OpenAI Agent Builder to n8n (MCP Integration)

AI Course of the Day

AI Coding Masterclass: From Beginner to Expert in 90 Minutes

Hands-on “vibe coding” course using top AI coding models to build websites, prototypes, web apps, games, and mobile apps—even with little or no coding. The presenter demos three models—Anthropic Claude/Claude Code, OpenAI GPT/Codex, and Google Gemini/Nano‑banana image—in a live editor workflow.