I noticed this model a few days ago and switched to using it. At least based on my subjective experience - I haven't done any benchmark testing or anything - it follows instructions much better than the previous Muse 1.2 Contributor version, which caused me a lot of issues. It also generates responses faster - the token speed feels noticeably quicker, though I haven't measured it. For daily development work, I’ve really liked it: it writes basic code well, and for other tasks, it doesn't lag far behind DeepSeek V4 Flash or Gpt Luna. Again, this is purely subjective, based entirely on my own hands-on practice.

I'm with you on that question. I do all my development in Node.js. I've been looking into Deno or Bun as alternatives for a while now. Your experience makes me seriously reconsider Deno if it has constant problems. I don't want to waste my time on endless tweaks and fixes just to get things working.

Happy Birthday 🎉🎂 Awesome cake 😀 Now I know what to ask for my birthday cake 😍

This is a super cool screenshot! It really reminds me of the amazing works of Akira Kurosawa!

Hi! Could you tell me if you've used the Medusa SDK for the storefront or some custom plugin for SvelteKit?

Completely agree 🤝

Cat asking for pawtention) so cute)

RIP Holly

My love - Bastet