Nvidia’s powerful open AI model play

Good morning, AI enthusiasts. The corporate selling the shovels within the AI gold rush just made an enormous move to begin mining, too.

With its recent powerful (and fully open) Nemotron 3 models, Nvidia is giving Western developers the competitive option they’ve been missing — while ensuring they stay hooked on its hardware in the method.

In today’s AI rundown:

• Nvidia releases Nemotron 3 open models for agentic AI

• AI reasoning models smash CFA exams

• Design higher web sites with Cursor’s recent editor

• Perplexity study shows how users are using AI agents

• 4 recent AI tools, community workflows, and more

LATEST DEVELOPMENTS

NVIDIA

🤖 Nvidia releases Nemotron 3 open models for agentic AI

The Rundown: Nvidia just introduced its Nemotron 3, a family of open models designed specifically for constructing multi-agent AI systems — marking the chipmaker’s most vital push yet into frontier model development.

The small print:

• The lineup spans three sizes: Nano (30B), Super (100B), and Ultra (500B), with Nano available now and the larger versions coming in 2026.

• Nano beats similar-sized models like Qwen3-30B on coding and instruction-following benchmarks, while also generating responses over 3x faster.

• Unlike most closed U.S. rivals, Nvidia is publishing training data, fine-tuning tools, and reinforcement learning environments alongside the models.

• The chipmaker lists Cursor, Perplexity, ServiceNow, and CrowdStrike as early adopters across coding, search, enterprise automation, and cybersecurity.

Why it matters: Closed U.S. labs are increasingly creating their very own silicon, while Chinese leaders are dominating open model usage globally. Nvidia releasing strong, completely open models gives Western developers a competitive open-source option, and likewise keeps them constructing on Nvidia’s hardware in the method.

TOGETHER WITH HUBSPOT

🧠 100+ ChatGPT prompts to revolutionize your workflow

The Rundown: HubSpot’s free, comprehensive “How you can Use ChatGPT at Work” guide provides 100+ ready-to-use prompts to assist professionals boost efficiency and adopt AI-driven workflows.

Inside, you’ll find:

• A fast crash course to master ChatGPT in under half-hour

• Practical industry use cases to spark real-world inspiration

• 100+ prompts to streamline tasks and speed up productivity

• Expert tricks to tackle common AI roadblocks with confidence

Get your free copy and join 10,000+ professionals leveling up with AI.

AI & FINANCE

🤑 AI reasoning models smash CFA exams

The Rundown: A brand new study just found that six leading AI models now pass all three levels of the Chartered Financial Analyst (CFA) certification exams, with Gemini 3.0 Pro scoring a record high of 97.6% on Level 1.

The small print:

• Researchers tested GPT-5, Gemini 3.0 Pro, Claude Opus 4.1, Grok 4, and DeepSeek-V3.1 across 980 questions spanning all exam tiers.

• GPT-5 topped Level II at 94.3%, while Gemini 3.0 Pro dominated essentially the most difficult constructed-response section with 92%.

• In 2023, GPT 3.5 failed the primary two levels, and GPT-4 passed only Level I — with the leap to near-perfect scores taking roughly 24 months.

• An NYU study in September also showed models passing all three levels, but with scores within the 70s vs. the near-perfect numbers of current frontier systems.

Why it matters: Acing a standardized test and handling day by day demands of economic evaluation are still very various things, however the speed of improvement on these exams is wild — and models mastering finance knowledge could shift the occupation’s value toward human skills like client judgment and relationship management.

AI TRAINING

🎨 Design higher web sites with Cursor’s recent editor

The Rundown: Learn tips on how to quickly arrange and use Cursor’s recent visual design editor to refine your frontend design without having to modify backwards and forwards with a design tool like Figma.

Step-by-step:

1. Open a brand new project in Cursor with an HTML and CSS file (latest update required) — ask Cursor agent to construct an easy index.html + styles.css or use a template

2. Install the live server extension: hit CMD+Shift+P, search “Open with Live Server,” then copy the URL and paste it into Cursor browser (CMD+Shift+B)

3. Toggle on the element selector, click any element to edit properties within the Design pane, or tell the agent what changes you would like within the lively chat

4. Hit Apply for the agent to make changes — click “Keep” or “Keep All” to save lots of (the agent updates classes mechanically so changes apply to all matching elements)

Pro tip: Make certain you’re saving your progress with Git as you go. Git will make it easier to roll back any unwanted style changes, and you may even ask the agent to put in writing Git commits for you.

PRESENTED BY TRIPLE WHALE

🎁 $2.9B in BFCM AI data is in

The Rundown: Triple Whale tracked $2.9B across 50K brands during this 12 months’s BFCM (19.7% of all Shopify sales), with Moby AI handling over 24K queries like “Which campaigns should I scale?” and “What should my spend distribution appear like tomorrow to hit $600K?”  — providing actionable answers for users in seconds.

Download the BFCM Breakdown Report back to uncover:

• Platform-by-platform performance across Meta, Google, TikTok, Amazon & AppLovin

• Industry benchmarks across 15 industries (Apparel led with 36% of revenue)

• The AI strategies top-performing brands used to optimize in real time

Get the report.

AI RESEARCH

📊 Perplexity study shows how users are using AI agents

The Rundown: Perplexity and Harvard just published a study analyzing how users of the corporate’s Comet browser are utilizing AI agents, finding usage heavily centered on cognitive work and research over automating easy tasks.

The small print:

• Researchers analyzed lots of of hundreds of thousands of anonymized queries from Perplexity’s Comet browser, which launched in July.

• Over half of queries involved research or workflow management, with common tasks including summarization, document editing, and coursework help.

• Tech employees, academics, marketers, and finance professionals generated the majority of activity, with adoption correlating to GDP and education levels.

• Behavior shifted over time, with users who began with casual queries like travel planning often migrating toward heavier knowledge work in a while.

Why it matters: Through the AI agent rise, we’ve ceaselessly seen use cases like booking flights/tickets, ordering groceries, and handling mundane tasks. But Perplexity’s research (which can skew in another way from ChatGPT or other assistants) shows users gravitating toward deeper work uses, not personal life conveniences.

QUICK HITS

🛠️ Trending AI Tools

• 📊 Quadratic – The AI spreadsheet that does your be just right for you. Turn raw data into insights and visuals without the pain of old tools*

• 📚 Gemini Deep Research Agent – Google’s SOTA agent for long-running context gathering and synthesis tasks

• 🤖 Nemotron 3 – Nvidia’s recent family of open-source models designed to construct multi-agent systems

• 🚀 Manus 1.6 – Recent performance upgrades for complex agentic work

*Sponsored Listing

📰 Every thing else in AI today

Merriam-Webster named ‘Slop’ as its 2025 Word of the 12 months, defined as “digital content of low quality that’s produced often in quantity via AI.”

The U.S. Office of Personnel Management introduced “Tech Force,” a push to recruit 1K early-career AI and software employees into federal government roles.

Manus released version 1.6 of its AI agent platform, which introduces mobile app development and a visible design editor alongside performance improvements.

Klarna launched the Agentic Product Protocol, an open standard that offers AI assistants access to over 100M products across merchants.

AI2 released Olmo 3.1, an upgraded version of its open-source model family that the lab claims is the “strongest fully open reasoning model.”

COMMUNITY

🤝 Community AI workflows

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today’s workflow comes from reader Vicky F. in Phoenix, AZ:

“I’m using AI as a research partner to reconstruct the complex lifetime of certainly one of my ancestors. ChatGPT has helped me locate and analyze historical records, draft emails to libraries and state agencies, and uncover recent leads….We’ve built timelines, tested genealogical theories, compared maps and newspaper accounts, interpreted probate language, and tracked migration patterns. Now we’re co-writing a full narrative of her life.”

How do you employ AI? Tell us here.

🎓 Highlights: News, Guides & Events

• Read our last AI newsletter: Gemini turns headphones into translators

• Read our last Tech newsletter: SpaceX preps monster $1.5T IPO

• Read our last Robotics newsletter: 1X’s home humanoid gets a factory job

• Today’s AI tool guide: Design web sites with Cursor’s visual design editor

• Watch our last live workshop: AI Automations Made Easy

See you soon,

Rowan, Joey, Zach, Shubham, and Jennifer — the humans behind The Rundown