Hello friends!
Welcome to another SLOTH BYTES. I hope you’re having a goofy week.
For the people with low attention spans:
Shopify bought Tailwind Labs - Yup they sold out. I would too.
CursorBench 4.0: Cursor upgraded its internal model leaderboard. Looks a bit more realistic? I don't know.
React 19.3: View Transitions are stable. LETS GOOOO.
Expo Modules 2.0: no more DSL, just write Swift like a normal person
Cognition's SWE-2: new AI model with good performance at a fraction of the price. The AI price war continues.
Latest Posts
If you're new, or not yet a subscriber, or just plain missed it, here are some of our recent editions:

Build. Break. Fix. Learn.
KodeKloud gives you 1,280+ hands-on labs where you provision Kubernetes clusters, write Terraform configs, build CI/CD pipelines, configure Linux systems, containerize apps with Docker, automate with Ansible, and manage Git workflows.
78+ playgrounds let you experiment freely in sandbox AWS environments, Kubernetes clusters, and CI/CD systems without risk.
190+ courses across DevOps, Cloud, and AI pair theory with hands-on labs at every step.
KodeKloud Engineer and 100 Day Challenges provide real-world job scenarios with automated grading that confirms your solutions work.
Stuck? The 55,000+ member Discord community connects you with peers and instructors ready to help.
Every lab runs in a live environment. You deploy, you troubleshoot, you learn. No videos without context. No simulations. The kind of practice that actually builds confidence because you've done real work, not watched someone else do it.
Domain Names + Web and Email Hosting You Need
Still paying GoDaddy or Namecheap prices? Porkbun sells most domains at cost for low, transparent registration and renewal pricing with no nonsense. Get free features like WHOIS privacy and SSL certificates, plus real human support 24/7, 365 days a year. Save $1 on your next domain name now.

Shopify just bought Tailwind
Looks like Tailwind Labs is joining Shopify. This is honestly a big deal.
Tailwind gets installed over 110 million times a week and styles sites like ChatGPT, X, Cloudflare, Reddit, and (of course) Shopify itself. Probably styles your projects too.
WHY WOULD THEY DO THIS???
Adam Wathan (creator of tailwind btw) says it’s to give tailwind “a stable long-term home” which is 100% reasonable given its situation in this AI era:

Yeah… It was rough.
Plus they were also one of the first companies to start building with it.
What stays the same: Tailwind CSS and the rest of their projects will remain MIT, and still be led by the same team.
What's changing: Well… they no longer need to try to grow the business around Tailwind now that they have sugar daddy Shopify funding them. This however means they’re closing sign ups for Tailwind Plus and ui.sh. If you already owned it, you keep your access. They'll instead be stepping back to focus on Tailwind CSS itself. Which is great!
Cursor upgraded CursorBench to 4.0
Cursor has upgraded their benchmark. I’ll just steal what Lee Rob said:
“It includes new tasks for how well models follow instructions, work on challenging projects over time, and is more difficult than before (so all models score lower).”
What data does it use?: Real coding sessions inside the Cursor IDE (bug fixes, refactors, multi-file changes), not scraped GitHub issues
Why Cursor cares: they use it internally to help decide which model becomes your default in the editor
The catch: it only measures a model running inside Cursor's own agent harness. Swap the harness and the same model can land somewhere completely different on other benchmarks
React 19.3: View Transitions are now stable!
React 19.3 landed on npm, and view transitions have graduated from experimental to stable. LETS GOOOO. It only took 3 years, but hey who cares.
But besides that there are 4 other new features:
Fragment Refs: As the name says you can now add refs to fragments. Doing this creates a FragmentInstance object.
browser: an API that lets you mark a component as browser-only during server rendering.Trusted Types support: React now integrates the browser’s Trusted Types API. It’s a security feature that helps prevent DOM-based XSS attacks.
<Context>can now be rendered directly in Server Components
Expo Modules 2.0: just write Swift bro
Expo Modules 2.0 is out, and it's about time. They're ditching the old result-builder DSL for something way more familiar: a native module is now just an @ExpoModule class with @JS on whatever methods and properties you want exposed. No more fighting a DSL only three people on the internet ever fully understood.
Sync vs async: comes straight from Swift's own
asynckeyword. Records, shared objects, and events all follow the same patternMigration: 1.0 and 2.0 can coexist, so you migrate one function at a time instead of a big-bang rewrite (there's even an
expo-migrate-moduleskill for agents)Availability: live now on iOS in SDK 57 (experimental macros), SDK 58 has the documented beta, Android is still cooking, views and generated TypeScript are next
Speed: Expo's own numbers put the
@JSpath at 2.5 to 5.6x quicker than 1.0 for sync calls on an iPhone 16 Pro. Not bad for a DSL's retirement party.
Cognition's SWE-2 model is out

Another week, another "we're basically the best model, just cheaper" launch.
SWE-2 is Cognition's (the Devin folks) most advanced coding model yet, and it scores 50.0% on FrontierCode 1.1 Main, which is their own benchmark for whether a human maintainer would actually merge an AI-written PR.
That's within a single point of Fable 5.1, at 64% less cost. It also lands a few points behind GPT-6 Astra for roughly a quarter of the price.
Once again, keep in mind these models can be trained to do well on benchmarks without necessarily performing as well in real-world scenarios. At this point though? This model is definitely good for fixing small bugs and creating new features if you’re specific.
What it's built on: it's post-trained on Moonshot AI's Kimi K3, a 2.8 trillion parameter open model.
Where to get it: live now in Devin Desktop and CLI, rolling out to Devin Web and Fusion. No public API, pricing, or open weights yet.
Special limited time deal: If you get a pro plan, they’re actually giving you UNLIMITED use of the model until October. So if you wanna vibe hard, might be worth it.
NOT SPONSORED BTW I SWEAR.

The Pulse: tech companies move to open AI models - AI is still expensive, so companies are optimizing their usage, check out some of their strategies so you can save money too.
How My App Is Doing (Kind of stuck) - My homie Chris gave an update on his calorie tracking app, great video if you’re interested in knowing what it’s like developing a mobile app with actual users.
Portal by Spotify - cut their Claude Code token usage by 90%, worth a read if you're burning through tokens fast.
Frontier engineering - Kiro's ten principles for actually working well with AI coding agents.

gpu-lexer - a tiny GPU-powered syntax highlighter from Vercel that works on any language automatically, fast but not flawless.
drawably - a fun hand-drawn UI library with zero dependencies, worth messing around with.
Fresh - a terminal code editor that feels like VS Code with zero setup, keybindings, mouse support, autocomplete, and split panes included.
OpenAI's Agents API - run coding agents in the cloud on the same engine behind Codex, no babysitting long sessions yourself.
Extend UI - React components for PDF, DOCX, XLSX, and CSV viewers, handy if you want your projects to interact with documents.

That’s all from me!
Have a great week, be safe, make good choices, and have fun coding.
If I made a mistake or you have any questions, feel free to comment below or reply to the email!
See you all next week.
Want to advertise in Sloth Bytes?
If your company is interested in reaching an audience of developers and programming enthusiasts, you may want to advertise with us here.








