Welcome to Jumble, your go-to source for AI news updates. This week, OpenAI split its flagship into three fresh models, while Elon Musk's SpaceXAI fired back with a coding model priced to undercut just about everyone. Let’s dive in ⬇️
In today’s newsletter:
☀️ OpenAI's flagship splits into three models
🚀 SpaceXAI answers with a half-price coder
⚖️ Whistleblower drags a famous hospital to court
📚 An Ivy League class gets caught leaning on AI
🥊 Weekly Challenge: Grok 4.5 versus the new Sol
🧠 OpenAI Launches the GPT-5.6 Family
OpenAI just launched the GPT-5.6 family, led by a new flagship named Sol. It claims state of the art results across coding, science, and security while burning fewer tokens for the same work.
What are your thoughts on GPT-5.6?
⚡ Three Models, One Family
Alongside Sol sit Terra for everyday work and Luna, the cheapest of the trio. The pitch is to match the model to the job instead of paying flagship prices for simple tasks.
🔬 Built For the Hard Stuff
Sol targets full knowledge work and pushes harder on cybersecurity and science than past models. OpenAI says it beats both its own older systems and rival frontier models.
🏷️ Where You Can Try It
The whole family is generally available now after a limited preview, and all three already show up inside popular coding tools. The whole story leans on more intelligence per dollar.
🛰️ SpaceXAI Drops Grok 4.5 to Undercut Everyone
Elon Musk's SpaceXAI released Grok 4.5, its first model built specifically for coding and autonomous agents. It is the first real product from the company's sixty billion dollar Cursor acquisition.
🧑💻 Trained Beside Cursor
The model was trained alongside Cursor and aimed at real engineering work rather than benchmark bragging rights. SpaceXAI calls it its strongest model ever.
💸 The Price Is the Pitch
Grok 4.5 lands at roughly half the price of comparable rivals. The bet is that developers care more about speed and cost than leaderboard scores.
Weekly Scoop 🍦
⚔️ Weekly Challenge: Grok 4.5 vs the New Sol
Challenge: Two of the biggest models of the year dropped in the same week. This week, put Grok 4.5 and ChatGPT's new Sol model head to head on your own work and see who actually wins.
Here's what to do:
📋 Step 1: Pick one real task Grab something you actually do, like drafting an email, cleaning up a spreadsheet, or fixing a bit of code. Keep it specific and real.
⚔️ Step 2: Run it in both Paste the exact same prompt into ChatGPT with the 5.6 Sol model, then into Grok 4.5 over at grok.com. Change absolutely nothing between the two.
🏆 Step 3: Crown a winner Compare the two answers for accuracy, tone, and how much cleanup each one needed. Whichever wins that task earns your subscription for it.
Is OpenAI's three model split the future, or is Musk's cheaper coding bet about to steal the whole show? See you next time! 🚀
Stay informed, stay curious, and stay ahead with Jumble!
Zoe from Jumble
