Meet Gemma 4 12B!
A unified, encoder-free multimodal model designed to bring high-performance intelligence directly to your laptop, and released under an Apache 2.0 license.
Bridging the gap between edge efficiency and advanced reasoning. Here is whatโs new with Gemma 4 12B: ๐
๐ Google is finally giving us what we've been asking for:
๐ AI performance reporting in Search Console
๐ซ A toggle to block content from AI Overviews & AI Mode
The catch? No click data. ๐ซฅ
And it's UK-only ๐ฌ๐ง for now.
Better than nothing, though. Details in comments.
@Nixous@RiseOfAlberta And in this world, when all your eggs are in one basket (the US), how do you fare if they elect a President who decides that climate change is a thing and should cut back on fossil fuel use? How about with 20 or 30% tariffs?
BC is a have province. We'll be OK.
๐จ SMX Advanced starts TOMORROW in Boston!
Join the brightest minds in SEO, PPC, AI, and analytics for advanced sessions, networking, meetups, and SMX After Dark. Tickets are still available, but time is almost up.
Register now: https://t.co/BqtwMx455L
@kennyhyder We all use IDEs and/or AI-IDEs (my preferred is Cursor), and I'm working in Railway at present on a larger-scale app I'm building, but sometimes you need a quick answer to a simple question, or you're on your phone. For those, you use the UI.
I'm constantly bouncing between Gemini, ChatGPT and Claude in the UI. I don't really have a formal "use model X on task Y" list, so I have to wonder how I intuitively select which will work better for whatever task I'm on.
And am I right?
@kennyhyder So tell us, oh wise one, what is the best UI-based LLM to use when one is looking for a quick answer or brainstorming?
I really want to catch up to where you must be.
Thanks in advance.
I'm looking forward to taking the stage at @smx in later this week (Thursday, June 4 at 1PM ET to be specific)!
You'll walk away with strategies you can use and even a couple Colabs to help you explore how to improve your visibility in ChatGPT and AI Overviews.
See you there!
The rumors are true ๐
Proud of the CoreWeave engineering team and our partners (like @Dell) who are the first to have a fully working @nvidia Vera Rubin NVL72, achieving yet another major milestone in bringing next-generation AI infrastructure online.
Stay tuned, there is so much more to come...
I have to give a BIG BIG shout out to the fine folks at @ScrunchAI for the great work they did on their MCP Server.
I've been working in it for the last few hours, and it's a really strong value-add to their product, helping me surface insights far faster and more intuitively.
The W&B MCP server is officially LIVE!
Coding agents could always read your code. Now they can read your experiments, monitor training, and drive their own research loops.
20 tools, hosted on every W&B deployment, plugs into Claude Code, Cursor, Codex, Gemini-CLI, and LeChat.
Is Gemini really SO bad at coding, that when asked to brainstorm ideas for a colab example, it suggests old versions of Claude and GPT as superior if I was to vibe-code? :D
@wandb Planning a GEO tracking tool that uses W&B to store the model outputs and key metrics.
Once that's done, my next project gets really interesting. :)
Introducing Gemini Omni ๐ฎ........ Omni is our new model that can create anything from any input โ starting with video (think Nano Banana but for video). Available in the Gemini App, Flow, and YouTube, with API support coming soon!
@CyrusShepard The doc read like they prompted Gemini with:
"Knowing the stock advice we've given webmasters on SEO for 20 years, adapt that to generative engines and format it for our blog."
Predictable and pretty much useless "advice". :(
The Gemini Pro model is rumored to be a GPT 5.5 level coding model ๐ง
The catch - it will be more than 50% cheaper at $12/1M output token
Gemini will take the lead over both GPT 5.5. and Opus 4.7 on the price-performance curve