hn.today

Mistral Large 4, sharing AI math progress, and Meta's Muse security issues

First things first

Some wild AI developments are shaking things up, from models that can design hardware to new tools for decision making. Privacy and security keep making headlines, especially with some big firms slipping up.

Plus, the tech world is grappling with burnout and legal drama while pushing boundaries with open-source projects and clever integrations. It’s a crazy, fast-moving scene out there.

The top 6

Mistral Large 4

Mistral Large 4

Mistral Large 4 is a multimodal AI model with 49 billion active parameters and a vision encoder, designed for general-purpose tasks. It offers features like structured outputs, function calling, and document Q&A, with a public preview available since October 2026.

1840 points · 1111 comments · docs.mistral.ai

Mistral Large 4: "Le Chonk"

Mistral Large 4: "Le Chonk"

Mistral Large 4, also known as Le Chonk, is a 1 trillion-parameter multimodal AI model that delivers state-of-the-art performance in coding, cybersecurity, finance, and law. It is trained and hosted in Europe, emphasizing AI sovereignty and open weights for enterprise control.

519 points · 5 comments · mistral.ai

Support our sponsor

Your Newsletter Deserves Better Than a Free Plan

Your Newsletter Deserves Better Than a Free Plan

beehiiv handles the unglamorous parts of running a newsletter - deliverability, referrals, ad revenue, subscriber growth - so you can spend your time writing. Built by people who scaled Morning Brew. Free to start, no card required. 14-days free, then save 20% for 3 months.

beehiiv.com

Quote of the day

The greatest danger for most of us is not that our aim is too high and we miss it, but that it is too low and we reach it.

Michelangelo

Also big today

Pass it on

Enjoying the digest?

Forward it to someone who'd like it, or send them the link to subscribe.

Share hn.today

One for the road

Did you hear about the circus fire?

It was in tents.

Worth a look

EmbeddingGemma 2

EmbeddingGemma 2

EmbeddingGemma 2 is a new multimodal embedding model developed by Google, capable of handling both text and vision data. It offers a moderate size of 270 million parameters for text and 440 million for combined text and vision, aiming to improve how large language models and AI agents work.

342 points · 35 comments · blog.google

Song of the day

This Is How We Do It

Montell Jordan;Wino

Deep cuts

The Early History of Smalltalk

The Early History of Smalltalk

Smalltalk was the first complete realization of a new paradigm in computing focused on object-oriented design, user interfaces, and end-user literacy. It emerged from sixties ideas about human-computer interaction and aimed to create a qualitative shift in how computers were used and understood.

120 points · 81 comments · worrydream.com

The daily digest

Today's best Hacker News stories, summarized and screenshotted, one email a day.