I created a minimal Nuxt 4 example integration with the Mastra agent framework: • Nuxt 4 app + API • Mastra agents + tools • Vercel for deployment github.com/Strift/nuxt-ma…

the Esports World Cup desk needs to suit up + start smiling wtf

🚨 GENG 🇰🇷 COMPLETED THE PERFECT SPLIT 26W-0L. Perfect Spring split. Their last series loss was 147 days ago 🤯

PoV: it's 2008 and you just ran home from school/work to play World of Warcraft Life is good

I took a one-month holiday, and now I'm unsure which LLM models to use for which task. It's both exciting... and annoying

🤣🤣🤣

Excited to finally get multi-modal embeddings out-of-the-box; this should make implementing image similarity search (à la Google Image) straightforward.

I used to think pretty highly of ARTE Info but the western propaganda bs lately is just too much 😕

One of my clients just edited one of my drafts and replaced all em dashes with regular dashes “nothing personal” they said 😭

📈 Why is ChatGPT sending more and more traffic to your website? Because Googling isn’t the only way to search anymore. Users now also discover content and products by talking with their favorite LLM client. Search and discovery happen through discussion, not keyword queries.

Is anyone using Anthropic API for personal use? The costs are just so damn high I just paid $3 for a request that would cost me *CENTS* with other providers 🤯 Am I doing something wrong?

🔗 scout:queue-import: Faster Indexing in Laravel Scout (by Nabil Hassen) nabilhassen.com/scoutqueue-imp…

The days of searching with text aren't over (yet), but there’s more to search than just typing keywords. 🔍 When shopping online, users might upload a picture and add ‘alternatives’. It’s now up to your product to figure out what they want. That’s where multi-modal search

Big shoutout to the Effect | TypeScript at Scale docs writer(s), it's the best I've read in a very long time

Hey Beeper, LINE Global when 👀

Outside of speed, I found both models to be good. Requires a little more prompting for research tasks than e.g. Gemini though Excited to try this out

Mistra 3.1 is a great contender with bigger models indeed. It's slowly become my go-to model for writing tasks. still have to see how it performs vs Gemini Flash for quick tasks requiring tool selection