OpenAI’s o3 and o4-mini arrive

Good morning, AI enthusiasts. OpenAI just released o3 and o4-mini, two recent reasoning models that president Greg Brockman called a “GPT-4 level qualitative step into the longer term.”

With o3 pushing SOTA across all benchmarks and supposedly capable of making recent scientific ideas, is that this the leap that finally puts the AI world at AGI’s doorstep?

In today’s AI rundown:

• OpenAI’s o3 and o4-mini, recent coding agent

• Copilot gets hands-on computer use

• Tips on how to run AI privately on your individual computer

• Claude gains autonomous research powers

• 4 recent AI tools & 4 job opportunities

LATEST DEVELOPMENTS

OPENAI

🤖 OpenAI releases o3 and o4-mini, recent coding agent

The Rundown: OpenAI just released o3 and o4-mini, its smartest reasoning models yet which can be now equipped with full agentic access to all ChatGPT tools and the power to “think with images” — alongside the launch of a brand new open-source coding agent.

The main points:

• OpenAI o3 is the brand new top-tier reasoner, pushing SOTA performance across coding, math, science, and multimodal benchmarks.

• o4-mini offers fast, cost-efficient reasoning, significantly outperforming previous mini models and even saturating benchmarks like AIME 2025 math.

• Each models can use and mix all tools inside ChatGPT (web search, Python, image generation, etc.) as a part of their problem-solving process.

• The models are also the primary to have the ability to “think with images”, integrating visual evaluation and manipulation directly into their chain of thought.

• Also launching is Codex CLI, an open-source coding agent that runs in users’ terminals and links reasoning models with coding tasks.

• President Greg Brockman said the discharge is a “GPT-4 level qualitative step into the longer term,” with the models capable of manufacturing novel scientific ideas.

Why it matters: Regardless of the bar for AGI is, it looks like the most recent SOTA models are getting close. While reasoners were already an enormous leap, equipping them with access to tools and multimodal capabilities has led to a category of models that’s creating recent ideas — seemingly taking us to Step 4 of OpenAI’s ladder of AI intelligence.

TOGETHER WITH AUGMENT CODE

🤖 The AI agent for skilled devs

The Rundown: Augment Code’s powerful AI coding agent meets skilled software developers exactly where they’re, delivering unmatched productivity and top-ranking performance.

With Augment Agent, you’ll experience:

• The No.1 ranked open-source agent on SWE-Bench by combining Claude 3.7 and o1

• Easy integration into Vim, JetBrains, and VS Code environments

• Compatibility with 100+ native and MCP tools

Enroll and begin constructing with Augment Agent.

MICROSOFT

🖥️ Copilot gets hands-on computer use

The Rundown: Microsoft just rolled out a brand new ‘computer use’ capability in Copilot Studio, enabling users and businesses to construct AI agents that may directly operate web sites and desktop applications.

The main points:

• The brand new feature allows agents to interact with graphical user interfaces (GUIs) by clicking buttons, choosing menus, and typing into fields.

• The method unlocks automation for tasks on systems lacking dedicated APIs, allowing agents to make use of apps similar to humans would.

• Computer Use also adapts in real-time to interface changes using built-in reasoning, mechanically fixing issues to maintain flows from breaking.

• All processing happens on Microsoft-hosted infrastructure, with enterprise data explicitly excluded from model training.

Why it matters: Copilot joins the likes of OpenAI and Anthropic’s Computer Use tools, marking one other step in AI’s agentic shift from chat windows into on a regular basis software. While it’s not the one UI automation tool, Microsoft users’ existing business workflows are an ideal use case to reap the benefits of one of these feature.

AI TRAINING

🤖 Tips on how to run AI privately on your individual computer

The Rundown: On this tutorial, you’ll learn how one can run powerful AI models directly on your individual computer for complete privacy, zero cost, and offline use—without sending data to external servers.

Step-by-step:

1. Select your platform by downloading Ollama or LM Studio based in your command-line or GUI interface preference.

2. Install the software and open it (each options can be found for Windows, Mac, and Linux).

3. Download an AI model that is suitable to your computer

4. Start chatting along with your AI using terminal commands in Ollama or the chat interface in LM Studio.

Pro tip: Match the model size to your computer’s capabilities; newer computers might have the ability to handle larger models (12-14B), while older ones should persist with smaller models (7B or less).

SPONSORED BY SALESFORCE

⏰ The countdown is on for Dreamforce 2025

The Rundown: Dreamforce 2025 — the world’s largest, most trusted AI event — returns on Oct. 14 with three days of next-level innovation and real-world inspiration you possibly can bring back to your small business.

Here’s what you’ll experience:

• AI agent constructing alongside product experts with Salesforce’s Agentforce

• 50+ visionary and product keynotes

• 1,200+ breakout sessions across every product, role, and industry

• 150+ hands-on trainings and 240+ peer roundtables

Join the list and get early access.

ANTHROPIC

🔍 Claude gains autonomous research powers

The Rundown: Anthropic just unveiled major upgrades to Claude, introducing autonomous research capabilities and Google Workspace integration to permit the assistant to look each the online and user files for answers with higher context.

The main points:

• The brand new Research feature can autonomously perform searches across the online and users’ connected work data, providing comprehensive, cited answers.

• A brand new Google Workspace integration lets Claude securely access user emails, calendars, and docs for context-aware assistance without manual uploads.

• Enterprise customers also get access to enhanced document cataloging, using RAG to look entire document repositories and lengthy files.

• Research is launching in beta for Max, Team, and Enterprise plans across the US, Japan, and Brazil, with Workspace integration available to all paid users.

Why it matters: Anthropic continues to maneuver at its own pace in terms of feature rollouts, giving Claude a “Deep Research” type feature well after the opposite major labs. But as we’ve seen with other rivals, the mix of web search, user data integration, and SOTA models can result in some extremely powerful results.

QUICK HITS

🛠️ Trending AI Tools

• 📽️ Veo 2 – Google’s SOTA video model, now available in Gemini App

• 🎥 KLING 2.0 Master – Recent video AI with improved prompt adherence

• ⚙️ Grok Studio – Canvas-like interface to collaborate with AI on docs and more

• 🔎 Embed 4 – Cohere’s recent multimodal search model for enterprises

💼 AI Job Opportunities

• 🤖 xAI – AI Engineer & Researcher

• 🛠️ Runway – Technical Customer Support

• 🧪 Anthropic – Research Engineer

• 🎨 Parloa – Lead Product Designer

📰 Every little thing else in AI today

OpenAI is reportedly in talks to amass coding platform Windsurf (formerly Codeium), in a deal value as much as $3B.

Microsoft researchers unveiled BitNet b1.58 2B4T, a brand new 1-bit AI model that matches the performance of larger models while running efficiently on CPUs.

Tencent introduced FireEdit, a brand new AI image editing system that uses region-aware vision language models to enable more precise, instruction-based image modifications.

Anthropic is reportedly preparing to launch a brand new “voice mode” for Claude with three distinct AI voices named Airy, Mellow, and Buttery this month.

OpenAI’s testing partner Metr published its evaluation of 3o and 4o-mini, noting an accelerated evaluation timeline that aligns with other reports of rushed safety testing.

Economist and writer Tyler Cowen said he believes o3 qualifies as AGI, questioning if April 16 can be the day the technology officially crossed the barrier.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop on Tuesday, April twenty second, at 3 PM EST with Matt Waters from Superhuman. By the tip of this workshop, you’ll have a completely optimized email system powered by AI, so you possibly can move 4x faster and eliminate inbox chaos.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

🤝 Share The Rundown, get rewards

We’ll all the time keep this text 100% free. To support our work, consider sharing The Rundown with your mates, and we’ll send you more free goodies.

See you soon,

Rowan, Joey, Zach, Alvaro, and Jason—The Rundown’s editorial team