They are optional animated companions for the Codex app: small floating overlays that sit on your screen while Codex works. The silly part is obvious. The useful part takes about five minutes to notice.

A Codex Pet can show whether Codex is running, waiting for input, or ready for review without making you switch back to the app every few minutes. If you already compare coding agents like Codex and Claude Code, this is the softer side of the same product question: how much does the tool help you stay in flow?

What Is a Codex Pet?

Codex Pets are animated companions built into the Codex app. You can turn one on from Settings > Appearance > Pets, type /pet in the composer, or use the command menu with Cmd + K on Mac or Ctrl + K on Windows.

Once active, the pet floats over your desktop and follows the state of the current Codex thread. OpenAI says the overlay can show the active thread, reflect whether Codex is running, waiting for input, or ready for review, and pair that state with a short progress prompt.

That means you can leave Codex visible in a small, ambient way while you do something else.

How to Turn On a Built-In Pet

If your Codex app supports Pets, the fastest way is to type this in the composer:

/pet

You can also open Settings > Appearance > Pets and choose from the built-in pets there. The same area lets you wake the pet, tuck it away, or refresh custom pets from your local Codex home.

If you prefer the keyboard route, open the command menu:

• Mac: Cmd + K
• Windows: Ctrl + K

Then search for the pet commands.

How to Hatch a Custom Codex Pet

You are not limited to the built-in pets. OpenAI also has a hatch-pet skill for creating a custom animated pet from a text concept, reference image, or both.

First, install the skill from inside Codex:

$skill-installer hatch-pet

Then reload skills. Open the command menu with Cmd + K or Ctrl + K, choose Force Reload Skills, and ask the skill to create a pet.

For example:

$hatch-pet create a tiny desk goblin

Or:

$hatch-pet turn my dog into a Codex pet

Or something looser:

$hatch-pet create a new pet inspired by my recent projects

The skill handles the pet-specific work: concept planning, animation rows, sprite sheet assembly, validation, previews, and packaging. The result is a Codex-compatible pet saved locally, which you can refresh and select from the Pets settings.

If you do not know what to hatch, Codex Pets is a small community gallery that can give you ideas for pet styles, themes, and prompts before you make your own.

What Makes a Good Pet Prompt?

Keep it simple. These pets are small, so clear silhouettes beat elaborate designs.

Good prompt ideas:

• a sleepy raccoon with a tiny laptop
• a calm duck that looks focused while Codex runs
• a blue-screen gremlin for debugging sessions
• a tiny robot with one glowing eye
• my dog as a pixel-style coding companion

Avoid asking for too many props, tiny labels, complex backgrounds, or realistic detail. A Codex Pet needs to read well as a small animated sprite, not as poster art.

If you use a reference image, describe what should carry over: color, shape, expression, accessory, or personality. Do not assume the skill will preserve every detail from a photo.

Why This Feature Works

Codex Pets sound like a joke until you treat them as ambient UI.

A coding agent spends a lot of time in states that are easy to miss: still working, waiting for permission, ready for review, stuck on input. A normal app badge or notification can tell you some of that, but a floating overlay keeps the state in your peripheral vision.

That is why the feature lands better than it should. It makes Codex feel present without requiring another panel, dashboard, or notification stream. If you want the broader OpenAI tools view, this older roundup of useful ChatGPT tools is a decent snapshot of how quickly the ecosystem keeps shifting.

Should You Try It?

If you use the Codex app, yes. Turn on a built-in pet first and see whether the overlay helps or annoys you. If it helps, install hatch-pet and make something that fits your workspace.

It will not write better code for you. It will not replace good review habits. But it can make long Codex sessions easier to track, and it gives the app a little more personality without getting in the way.

The post Codex Pets: What They Are and How to Hatch Your Own appeared first on Hongkiat.

You open a blank canvas, try a few hex codes, hate all of them, then end up nudging saturation sliders for twenty minutes like that somehow counts as design progress.

I have been there more times than I want to admit.

There are plenty of websites that can help you generate palettes, browse combinations, test contrast, and build gradients without fighting your tools. If you want to understand why some combinations work and others do not, this guide to the basics behind color theory is a useful companion. Here are five worth bookmarking.

1. Coolors

Coolors is still one of the fastest ways to get unstuck.

Its main draw is the palette generator. Tap the spacebar, and it keeps serving new combinations until something clicks. That alone makes it useful when you need quick inspiration, but Coolors goes further than that.

It also includes palette exploration, an image picker, a contrast checker, a palette visualizer, gradient tools, and even Tailwind-focused color helpers. If you want one site that can take you from rough inspiration to something you can actually test in a UI, this is probably the most complete one on the list. For a broader view of what’s available, see our roundup of free color tools for web designers.

2. Grabient

Grabient is the one I would open when the brief calls for gradients first, not flat palettes.

It is built around generating CSS gradients with a clean interface and quick export options. The site also doubles as a palette finder, so it is not limited to flashy background blends, but gradients are clearly the star here.

If you build landing pages, hero sections, or marketing visuals and want gradients you can tweak without wrestling with CSS by hand, Grabient is a very handy shortcut.

3. ColorHub

ColorHub feels more product-minded than a lot of palette tools.

Instead of stopping at color generation, it tries to help you preview palettes in more realistic layouts. Its newer versions also added starter kits, palette previews on landing-page style templates, a contrast checker, favorites, and curated palette libraries.

That makes it especially useful if you are not just hunting for a nice palette, but trying to figure out whether that palette will still look good once it hits an actual interface.

4. Color Hunt

Color Hunt is simple in the best way.

It is basically a big, browsable collection of hand-picked color palettes, organized by mood, theme, and style. Pastel, retro, dark, wedding, food, sunset, neon. You can drift through it for five minutes and usually come away with something usable.

I like it because it does not overcomplicate the job. Sometimes you do not want a full palette lab. You just want to browse good combinations until one feels right.

That is exactly what Color Hunt is good at.

5. Gradients by Colorion

If all you need is a gradient and the CSS for it, Colorion keeps things nice and direct.

The site offers a hand-curated collection of background gradients, lets you change direction, and gives you the generated CSS immediately. No long setup. No extra friction. Just browse, tweak, copy, and move on.

It is especially useful for quick web design work when you want a softer background, a more interesting callout section, or a color blend that looks more polished than a flat fill.

Which One Should You Start With?

If you want the short version:

• Use Coolors when you want the broadest toolbox.
• Use Grabient when gradients are the main job.
• Use ColorHub when you want to preview palettes in something closer to a real UI.
• Use Color Hunt when you just want to browse good color combinations fast.
• Use Colorion when you want ready-made gradients and CSS with almost no effort.

Most of us do not have a color problem. We have a workflow problem.

The right tool does not magically give you taste, but it does help you explore faster, compare better, and stop second-guessing every hex code. That alone is a good reason to keep these five bookmarked. If you are starting a new project and still working through what colors to use, how to choose a color scheme for your website goes deeper into that decision.

The post 5 Color Palette Generator Websites Worth Using appeared first on Hongkiat.

Binky is a native macOS app for cleaning up those fussy folders. It can sort a folder on demand, watch folders continuously, and move files according to rules you control. It waits for files to finish arriving first, then routes them into folders like Images, PDFs, Media, Documents, Archives, Apps, Screenshots, and Misc.

The point is not clever file management. It is predictable file management.

Quick Sort Handles One-Off Cleanup

Binky’s Quick Sort is for the folder you want to clean right now. Pick an inbox, run the sort, and Binky sweeps through the files in one pass.

Downloads is the default target, but it is not the only one. You can use it on any folder that has become a dumping ground: Desktop, a shared folder, a client upload folder, a screenshots folder, or wherever your Mac stores chaos. If your issue is less about sorting and more about Finder itself feeling limited, a utility like XtraFinder sits in the same Mac productivity lane.

By default, sorted files land in folders such as:

• Images/
• PDFs/
• Media/
• Documents/
• Archives/
• Apps/
• Screenshots/
• Misc/
• Review/

Binky waits for files to settle before moving them, so it does not grab a download halfway through. It also uses collision-safe moves, so two files with the same name do not overwrite each other.

Anything unknown or questionable goes to Review first. That is the part that makes the app feel safer than a blunt extension-based sorter. If Binky is not confident, it does not bury the file somewhere random.

Routines Watch Folders Continuously

Quick Sort is manual. Routines are automatic.

A Routine is a named watcher with its own source folder, rules, and optional Finder tags. You can have one Routine for Downloads, another for Desktop, another for a Dropbox inbox, and another for a screenshots dump. Each one can follow different routing logic.

That makes Binky more flexible than a simple Downloads cleaner. It can sit behind the places where files enter your workflow and keep them from becoming junk drawers.

Rules Give You Control

Binky can sort by file type, but it is not limited to extensions.

Rules can match by name, extension, file kind, size, date, origin, OCR or receipt hints, and Finder tag conditions. A rule can move a file, rename it, apply Finder tags, extract an archive, install from a DMG, trash a match, or fan files out by tag.

A few practical examples:

• send invoices and receipts into a finance folder
• keep client uploads away from general downloads
• route screenshots separately from normal images
• move project exports into the right project folder
• tag files during sorting so they are easier to scan in Finder

Rules run before the default sorted folders. So if you want receipts handled separately, Binky can catch them before they fall into a generic PDFs folder.

It Fits Into Mac Workflows

Binky is built with SwiftUI and AppKit, so it behaves like a Mac app rather than a web app wrapped in a window.

It supports Finder Quick Actions and Services, which means you can select files in Finder and run Sort with Binky directly. If you like tuning Finder actions yourself, this guide to adding items to the macOS context menu is a useful companion.

It also includes an Apple Shortcuts action called Sort Files, so you can pass file paths to the running app from your own automations. For a broader example of Mac automation, see this walkthrough on how to automate opening and positioning apps on Mac.

There is a binky command-line tool too, for people who want to preview sorts, run Routines from Terminal, or plug the same rules into scripts. The app, CLI, and Routines use the same preferences and routing logic, so you are not maintaining separate systems.

Binky also keeps history and batch summaries. You can see what moved, what was skipped, and what went to Review. Undo is available where macOS allows it.

What It Does Not Do

Binky moves files. It does not copy them. When a file is sorted, the original leaves the source folder and goes to its destination.

It also only watches folders while the app is running. If Binky is closed, new files stay where they are. Open it again or run Quick Sort, and it can clean up what accumulated.

There is no cloud account involved. Sorting happens on your Mac, and Binky does not upload your files.

The Bottom Line

Binky is a small Mac utility for a common problem: folders that turn into junk drawers because sorting files by hand is boring.

The app is more useful when you stop thinking of it as a Downloads cleaner and treat it as an inbox cleaner. Point it at any folder where files pile up, set a few rules if the defaults are not enough, and let Review catch anything uncertain.

Binky requires macOS 14 Sonoma or later. It is open source under the MIT license, and the 1.x line is free. You can download the DMG from binkyfiles.com and install it like a regular Mac app.

The post Binky Is a Free Mac App That Sorts Messy Folders for You appeared first on Hongkiat.

It helps users scan faster and saves you from drawing the same magnifying glass, hamburger menu, or notification bell over and over. If you want to cast a wider net, this list of icon search engines is a useful companion.

The harder part is finding libraries that are pleasant to browse, visually consistent, and broad enough that you do not run out of options halfway through a project.

These 11 cover a good spread: open-source defaults, premium systems, huge libraries, and a few more style-specific picks.

1. Lucide

Lucide is a clean open-source icon library with a crisp, minimal style that drops into modern web apps without fighting the rest of your interface.

It is more than a website with SVG downloads. It also has packages for React, Vue, Svelte, Solid, Astro, React Native, and more, which makes it practical for actual product builds rather than one-off mockups.

If you want lightweight outline icons with a consistent look, Lucide is easy to keep in rotation.

2. Iconsax

Iconsax targets designers who want range. It is more commercial than barebones open-source libraries, and the main draw is sheer volume. The site has tens of thousands of icons across multiple styles, framework support, a Figma plugin, and AI-assisted icon generation.

It works well when you need one library that covers dashboards, mobile UI, admin screens, and edge-case icon needs without switching visual styles halfway through.

3. The Noun Project

The Noun Project is less of a single icon pack and more of a huge icon marketplace. You get a massive, diverse collection of symbols made by many contributors, which helps when you need a very specific concept that most UI icon sets do not cover. The tradeoff is consistency, because not every icon will feel like it belongs in the same interface.

For presentations, editorial graphics, concept-driven visuals, or hard-to-find symbols, it is a useful place to look.

4. Iconic

Iconic keeps things straightforward. The site calls its icons pixel-perfect, works on a 24 by 24 grid with a 1.5px stroke, and splits the collection between free and pro sets. The free license is usable, which puts it ahead of icon libraries that bury the useful stuff behind pricing caveats.

The library covers interface-friendly categories and everyday UI shapes, with a style that sits comfortably in web products.

5. Hugeicons

Hugeicons is built around scale. The site positions itself as a giant icon system for both designers and developers, with framework packages, a Figma plugin, icon fonts, and enough breadth to cover almost any app category you can think of.

You pick this when you want scale, coverage, and a polished commercial ecosystem around the icons rather than something intimate or opinionated.

6. Pixelarticons

Most icon libraries aim for smooth and neutral. Pixelarticons goes in the opposite direction with a full pixel-art look. If you are designing something retro, game-like, playful, or deliberately low-fi, this fits better than forcing a generic outline icon pack into the wrong project.

The library has thousands of icons across multiple styles, with a free tier and a pro upgrade for the full set.

7. Nucleo

Nucleo is built around workflow as much as icon count. It combines a large premium library with its own app for browsing, customizing, and exporting icons, plus React packages for developers.

If you work on larger design systems or manage assets across teams, that extra layer is useful. Plenty of icon sites stop at “here are some SVGs.” Nucleo is built more like a full toolkit.

8. Iconoir

Iconoir is an open-source option that is modern, flexible, and genuinely free without the usual premium bait hanging over every click. Its style is clean and slightly expressive without becoming noisy. The project also supports multiple outputs, including SVG, font, React, React Native, Flutter, Figma, and Framer.

If you want a free library that still feels product-ready, Iconoir is easy to recommend.

9. Feather Icons

Feather is simple, open source, and very restrained. No marketplace pitch. No bloated interface. Just a clean collection of icons with adjustable size, stroke width, and color.

If your taste runs minimal and you do not need thousands of niche symbols, Feather is an easy default.

10. Phosphor Icons

Phosphor is a flexible icon family used widely in product and interface work. It is broad and designed to work across UI, diagrams, presentations, and general digital product design.

If Feather feels too sparse for you, Phosphor is often the next step up.

11. Pikaicons

Pikaicons is more playful than the flatter, more clinical icon libraries out there. The site pitches it as a modern icon library for web and mobile, with more than 5,000 icons, multiple styles, and a Figma-first workflow. That makes it useful for designers who live in Figma and want something with a bit more charm without losing consistency.

Pikaicons works well when you want your UI to look polished without making it feel sterile.

The post 11 Icon Pack Websites Designers Should Bookmark appeared first on Hongkiat.

If you enjoy this kind of thing, our earlier roundup of interesting websites you should bookmark goes in a different direction but scratches the same itch. This list stays practical: 11 websites worth bookmarking because they do something well.

iLoveIMG

iLoveIMG is one of the easiest places to start when you need to resize, compress, crop, convert, watermark, or clean up images without opening Photoshop. It packs a long list of image tools into a simple browser interface, and most of them work in bulk, which is where the real convenience kicks in.

What makes it useful is how little setup it asks from you. Drop in a few JPGs or PNGs, pick the action you need, and you are done. If you like collecting this kind of utility, these online tools for web design projects are a good companion read.

BookFinder

BookFinder is a price comparison engine for books. Search for a title, author, or ISBN, and it pulls listings from a large number of booksellers so you can compare prices for new, used, rare, and textbook editions in one place.

It is especially handy when you are chasing an out-of-print title, an older edition, or the cheapest possible copy of a book you only need for reference. Instead of bouncing across marketplaces one by one, you get a broader view of availability, condition, and pricing almost immediately.

Gapminder

Gapminder is one of the best websites for anyone who likes data but does not enjoy being buried in dry reports. Its mission is to challenge common misconceptions about the world using interactive charts, short explainers, and quizzes built around global trends.

You can use it to explore data on poverty, life expectancy, education, population, energy use, and more. Teachers, journalists, students, and curious readers will get the most out of it, but even a quick visit can reset a few assumptions about how the world is changing.

AsanConvert

AsanConvert does one very specific job that will make perfect sense to a very specific crowd: it converts CorelDRAW files across versions and exports them into formats such as PDF, AI, EPS, SVG, and PNG. If you have ever received a .cdr file that your version of CorelDRAW refuses to open, this site is aimed straight at that headache.

That narrow focus is exactly why it is useful. Instead of hunting for desktop workarounds or asking someone else to re-save the file for you, you can upload it and convert it in the browser. Not everyone will need this, but the people who do will probably bookmark it on the spot.

curlconverter

curlconverter takes a raw curl command and turns it into code for other languages and libraries. Paste in a request copied from your browser or terminal, then convert it into Python requests, JavaScript fetch, Node.js, Go, PHP, and plenty more.

For developers, this is a small but excellent shortcut. A copied curl request is often the fastest way to capture an API call, but not the nicest thing to drop into an app. curlconverter handles the translation step quickly, and the browser-side conversion is a nice touch too.

ProPublica Nonprofit Explorer

ProPublica’s Nonprofit Explorer lets you browse tax filings from millions of U.S. tax-exempt organizations. You can search by nonprofit name, person, city, keyword, or EIN, then dig into revenue, expenses, executive compensation, and full filing text.

This is a strong research tool for journalists, donors, job seekers, and anyone doing due diligence. If you want to know how a nonprofit is funded, how large it has become, or what its filings actually say, this site makes that process much easier than digging through scattered public records on your own.

archai

archai is a short adaptive assessment that tries to measure how ready you actually are for AI. According to the site, it takes about eight minutes and ends with a single result you can share.

What I like about the idea is that it does not present itself as another broad AI news or prompt library site. It feels more like a focused diagnostic. If you are curious where you stand, or you want a lightweight conversation starter for teams thinking about AI adoption, it is a neat little tool.

Ninite

Ninite has been around for years, and it is still one of the cleanest Windows utilities on the web. You pick the apps you want, download one custom installer, and Ninite handles the rest. No bundled junk, no endless Next buttons, no toolbars quietly sneaking in.

It is excellent for setting up a new PC or refreshing an old one. Browsers, messaging apps, utilities, runtimes, developer tools, media players, compression tools, and more are all there. For Windows users, it sits nicely alongside our list of free Windows tools you should install.

CEOExpress

CEOExpress feels like an old-school web portal, and that is exactly the charm. It collects links to business news, financial resources, research tools, travel info, weather, newspapers, and other executive-focused utilities in one dense homepage.

The design is unapologetically dated, but that almost helps. There is very little fluff between you and the links. If you like the idea of a start page that acts more like a control panel than a polished content experience, CEOExpress still does that job surprisingly well.

Music-Map

Music-Map is a music discovery tool built around one question: if you like this artist, who else should you listen to? Type in a band or musician, and it returns a visual map of nearby artists that listeners tend to associate with them.

It is refreshingly simple. No playlist bait, no recommendation essay, no account needed. Just type a name and follow the trail. If you are trying to branch out from an artist you already love, this is a fast and enjoyable way to do it.

Worldometer

Worldometer is a real-time statistics website that tracks numbers across population, health, food, energy, water, environment, economics, and more. It is best known for live counters and fast-glance global stats, which is why it keeps popping up whenever people want a quick sense of scale.

The appeal is obvious: the data is presented in a way that feels immediate. Open it and you are not digging through a report. You are looking at moving numbers, ranked tables, and simple topic pages that make large abstract quantities easier to grasp. It is one of those reference sites that works because it gets to the point quickly.

Final Thoughts

Most people use the same handful of websites every day and rarely look beyond them. That is why lists like this are still fun to build. The web gets a lot more interesting once you start collecting niche tools, odd utilities, and sites that do one job well.

And if you know a few more in this vein, send them over. These lists are never really finished.

The post 11 Useful Websites You Should Know appeared first on Hongkiat.

GhostTrack is a Python command-line tool that groups a few public data lookups into one menu. You can run it locally, pick an option, and get results without setting up anything complex. I spent some time with it to see what it actually does and where it falls short.

What you get with GhostTrack

After installing and running it, the menu gives you four options:

• IP Tracker
• Show Your IP
• Phone Number Tracker
• Username Tracker

The IP Tracker takes any address you throw at it and pulls country, city, region, ASN, ISP, timezone, and rough coordinates from ipwho.is. Running it against a test IP gave back readable output in seconds. To understand how much an IP can already reveal on its own, this breakdown of what your IP exposes is worth reading alongside it.

The Phone Number Tracker parses a number using Python’s phonenumbers library and returns region, carrier, timezone, formatting, and number type. The Username Tracker checks whether a handle shows up across a list of social platforms by hitting profile URLs and reading the response.

If username lookups are what you are after, Sherlock is worth comparing. It covers more platforms and scales larger, but the idea is similar.

This is not a live location tracker. It wraps public lookups and common Python libraries into something a bit more convenient to use.

What actually works

That does not make it useless.

If you are running beginner OSINT labs, security demos, or CTF prep and need one quick interface for basic recon, GhostTrack lowers the friction. One script, one menu, readable output without stitching together separate commands. That is the main appeal: convenience over capability.

The code itself is small and readable. For someone learning how these lookups work, it is a decent starting point to see how IP metadata, number parsing, and username checks are handled programmatically.

Where the pitch overstates what it does

The marketing leans on phrases like “track location” and “track mobile number,” which sound more dramatic than what actually happens when you use it.

The IP lookup gives approximate geolocation from an API, useful context but not live tracking. The phone lookup tells you carrier and region, not where someone is right now. The username check mostly reports whether a handle exists on a platform.

Going in expecting precision surveillance will disappoint you. Going in expecting a quick public-data aggregator makes the tool feel more honest.

Rough edges you will notice

The username checker flags HTTP 200 as a match, which is blunt. Some sites serve friendly pages even for accounts that do not exist, so false positives are likely. The phone-number tracker also defaults to Indonesia as the region, understandable given the author’s background, but not useful if you are working with numbers from elsewhere.

Beyond that, there is very little documentation, no validation notes, and no discussion of error cases. For a quick personal script this is fine, but it means you have to manually verify anything important.

Who it actually suits

GhostTrack makes most sense for:

• beginners exploring how OSINT lookups work
• hobbyists who want a menu instead of running separate commands
• quick demos showing what public metadata can reveal

It is less suitable for professional investigations, anything where evidence quality matters, or workflows that need clean reporting and real validation. It is not Maltego or a full recon platform. It is a small script behind a terminal menu.

The Privacy Angle

What stands out most after using it is not GhostTrack itself. It is how much metadata ordinary systems already expose. IP geolocation, carrier info, timezone hints, and username reuse patterns all add up. Even basic lookups can sketch a surprising amount of context when stacked together.

Defensively, that is a good reason to audit your own digital footprint. If you want to take that further, this guide on deleting your digital footprint is a practical next step.

The post How Good Is GhostTrack for IP Phone and Username Lookups appeared first on Hongkiat.

The problem is that some of Brave’s best features are buried in settings, while some of its more distracting ones are turned on by default. If you use Brave every day, a few quick changes can make it cleaner, faster, and less annoying.

Here are 10 settings and tweaks worth changing first.

1. Clean Up the New Tab Page

Brave’s new tab page can feel busier than it needs to be. By default, it pushes Brave Rewards, VPN promos, stats cards, and other extras that most people do not need every time they open a new tab.

Open a new tab, click the gear icon, and start trimming it down. You can turn off new tab ads, hide Brave-specific cards like Rewards, VPN, Talk, and News, and remove top sites if you prefer a blanker start.

While you are there, you can also change the search engine shown on the new tab page. If you want to replace Brave Search everywhere, head to Settings > Search engine and change the default for both normal and private windows. If privacy is the main reason you use Brave, you may also want to launch Brave in private mode by default.

2. Make Brave Shields More Aggressive

Brave already blocks a lot by default, but its content filters get much better once you dig into them.

Go to Settings > Shields > Content filters > Show full list and enable the filters you actually want. A few useful ones include:

• cookie notice blocker
• annoying distractions blocker
• AI suggestions blocker
• newsletter popup blocker
• YouTube Shorts blocker
• tracking URL blocker

Spend five minutes here and a lot of the web becomes noticeably less irritating. If you use Brave on your phone too, this goes nicely with other tips on blocking ads in Android.

3. Use Copy Clean Link When Sharing URLs

A lot of links now come stuffed with tracking parameters. You copy a URL, paste it somewhere else, and drag a pile of analytics baggage along with it.

Brave has a simple fix built in. Right-click a link and choose Copy Clean Link. It strips out a lot of that tracking junk before you share it.

It will not solve every case, especially shortened links, but it is still a smart little feature that more browsers should copy.

4. Turn Off Brave’s AI Features if You Do Not Want Them

Brave includes its own AI assistant, Leo. If you have no interest in browser-level AI, you can strip it out pretty quickly.

Go to Settings > Leo and disable the options there. Then right-click the toolbar icons you do not want and remove things like Leo, Brave Wallet, Brave Rewards, or Brave VPN from the address bar.

For a cleaner autocomplete experience, check Settings > Appearance and turn off the Leo suggestion there too.

5. Switch to Vertical Tabs

Vertical tabs make more sense than horizontal ones once your tab count gets even mildly embarrassing.

In Brave, go to Settings > Appearance > Tabs and enable Use vertical tabs. If you want a tighter layout, turn off Show title bar too. That removes one of the extra bars at the top and helps the browser feel less cramped.

On larger displays, this setup works much better. You can see more tab titles, waste less horizontal space, and avoid the tiny-tab mess that happens when too many pages are open.

6. Enable Memory Saver if You Hoard Tabs

If you are the kind of person who treats tabs like bookmarks, Memory Saver is worth turning on.

Brave can suspend inactive tabs to reduce memory usage and keep the browser from dragging your whole system down. You will find it under Settings > System. If there are sites you always want to keep alive, such as music players, chat apps, or dashboards, add them to the exception list.

This is a real quality-of-life feature if you keep dozens of tabs open. If that sounds like you, these tips for manage Chrome tabs and memory are useful too, even if you never touch Chrome again.

7. Use Brave’s Built-In Task Manager

When Brave starts feeling sluggish, the culprit is often one rogue tab, one bloated web app, or one extension going feral.

Press Shift + Esc to open Brave’s task manager and see what is eating CPU or memory. If the shortcut does not work, go to More Tools > Task Manager from the main menu.

This is one of those features people forget exists until the day a single tab starts torching their laptop.

8. Use Force Paste on Sites That Block Normal Paste

Some websites still block paste inside login forms or other fields, usually in the name of security and usually to everyone’s annoyance.

Brave includes a Force Paste option in the right-click menu for forms, which can bypass those blocks. It is a tiny feature, but a useful one if you deal with password managers or long strings of text all day.

If a site also blocks right-click entirely, you may still need an extension or workaround, but Force Paste handles the simpler cases well.

9. Use Brave on Mobile for a Better YouTube Experience

Brave’s mobile browser has a practical advantage over the regular YouTube app. It blocks ads, supports background playback, and can keep videos playing in picture-in-picture mode.

That gives you a chunk of the YouTube Premium experience without paying for YouTube Premium, especially on mobile where ad blocking is usually more limited.

If you already use Brave on desktop, this is one of the best reasons to install it on your phone too.

10. Force Dark Mode on Websites

A surprising number of sites still ignore dark mode. Brave has an experimental flag that can force dark themes across the web.

Type brave://flags/#enable-force-dark into the address bar, enable the option, and restart the browser.

Because this uses Brave’s own internal flag system, it is usually lighter than installing yet another extension just to darken websites. It will not render every site perfectly, but it is often good enough.

The Best Brave Setup Is Usually the Less Brave-y One

Brave is at its best when it gets out of the way.

The browser already has a strong foundation: Chromium compatibility, built-in blocking, decent privacy defaults, and a few genuinely useful tools. The bigger win is removing the stuff you do not want, then turning on the features that actually improve day-to-day browsing.

Do that, and Brave feels less like a browser packed with side projects and more like a lean Chrome alternative that respects your time a bit more.

The post 10 Brave Browser Settings Worth Changing First appeared first on Hongkiat.

In OpenAI’s official announcement, the company says the new system is faster, follows instructions better, preserves details more reliably across edits, renders text more cleanly, and now lives inside a dedicated Images experience in ChatGPT. It also says outputs can be generated up to four times faster than before.

That alone would make it a meaningful product update. The more interesting shift is how believable these images are starting to look, especially when the model is asked to generate screenshots, UI mockups, or desktop scenes that look like they came from a real device. If you want a quick refresher on how to create AI images in ChatGPT, this release pushes that workflow much further.

What the Launch Actually Adds

It can generate original images, edit uploaded photos more precisely, preserve details across multiple revisions, handle smaller and denser text better, and produce outputs in more aspect ratios. OpenAI also says the model is better at keeping people and objects consistent across edits, which helps when you are refining the same image over multiple passes.

This pushes the model closer to practical output, not just visual flair. The official examples point to mockups, explainers, social graphics, layouts, product visuals, and edits that hold together across iterations. That should feel familiar if you have already experimented with the ChatGPT AI image generator for more stylized or illustrative work.

If you want to browse examples yourself, OpenAI’s announcement and image generation docs are the best places to start.

The Screenshot Problem Just Got Worse

AI-generated screenshots used to be pretty bad.

The text usually gave it away first. Menu labels looked off, terminal output felt fake, browser tabs were messy, and interface spacing never quite matched what a real app or operating system would produce. Even when the image looked decent at first glance, it usually fell apart the moment you looked closer.

That is changing fast.

OpenAI is openly calling out better text rendering, and that is a huge deal for screenshots. Once labels, commands, notifications, browser chrome, and window layouts start looking clean, the usual tells start disappearing too.

Just look at the image below. It looks like a normal macOS Terminal window showing a Homebrew installation in progress. The command line layout feels plausible, the spacing looks natural, and the whole thing has the plain, unremarkable look of a real screenshot.

That is the shift. AI screenshots used to look like rough mockups pretending to be real. Now they can look close enough to pass in a post, a chat, or a support thread unless you stop and inspect them carefully.

Or look at the image below. It is not a screenshot exactly. It looks more like a photo of a phone showing a WeChat conversation, as if someone used another phone to take the picture. That extra layer makes it feel even more believable.

If the model can now generate cleaner UI text, stronger layout consistency, and more believable desktop or mobile compositions, then it is useful not just for interface mockups, but also for screenshots and screen photos that pass as documentation.

There Is a Real Upside Too

For legitimate work, this gets useful fast. Designers can mock up product screens without spending hours in Figma. Marketers can create cleaner campaign visuals. Developers can generate UI concepts, onboarding flows, and explainer graphics with less prompt wrestling than before. Anyone producing tutorials or documentation can prototype visual ideas much faster.

That is the tension. The same improvement that makes the tool more useful also makes it easier to abuse.

The post ChatGPT Image 2.0 Is Making Fake Screenshots More Convincing appeared first on Hongkiat.

This guide covers practical ways to debug a stopped Docker container, from checking logs and exit codes to preserving its state before you restart anything.

Check container status and logs

When a container stops, your first stop should always be the Docker logs. Docker keeps a record of everything that was written to stdout and stderr while the container was running.

First, let’s see all containers, including stopped ones:

docker ps -a

Look for your container in the list. Note its name or container ID.

Then, check its logs:

docker logs [container_name_or_id]

This often reveals the immediate cause. Maybe your application threw an exception, or a dependency failed to start. If the logs don’t show anything obvious, don’t worry. We have more tools.

Understanding exit codes

As you can see from the screenshot above, every Docker container exits with a code. Exit code 0 means success (the container completed its task). Any other number indicates an error. You can see the exit code in the docker ps -a output, or get it directly:

docker inspect [container_name_or_id] --format='{{.State.ExitCode}}'

Common exit codes:

If you see exit code 137, your container likely hit a memory limit and was killed by the system. Exit code 143 usually means something asked Docker to stop the container gracefully.

Before you try to bring the container back up, there is one mistake worth avoiding.

Don’t restart too soon

The usual instinct is to run docker start [container] or reach for Docker Compose. That can erase useful evidence, especially if the container was not configured with persistent storage. You could lose:

• Application logs that weren’t captured by Docker’s logging driver
• Temporary files created during the failed run
• Crash dumps or core files
• Database transaction logs (if running a database)
• Configuration changes made at runtime

So before you even think about restarting, you need to preserve the evidence.

Preserve logs before they disappear

Docker keeps logs for stopped containers, but there are limits. By default, Docker uses the “json-file” logging driver with no size limit, but in production, you might have log rotation or different drivers.

First, save the logs to a file immediately:

docker logs [container_name_or_id] > container_logs.txt

For containers with a lot of output, you might want to limit to the last N lines:

docker logs --tail 1000 [container_name_or_id] > recent_logs.txt

If you suspect the issue happened a while ago, you can include the timestamps:

docker logs --timestamps [container_name_or_id] | grep -i "error\|exception\|fail"

Save container filesystem state

When a container stops, its filesystem still exists unless it was started with --rm.

You can extract files from it using the following commands:

Copy files out before restarting

Use docker cp to extract important directories:

docker cp [container_name_or_id]:/var/log ./container_logs
docker cp [container_name_or_id]:/tmp ./container_tmp
docker cp [container_name_or_id]:/etc ./container_etc

Look for application-specific directories too. If you know your app writes to /app/logs or /data, copy those.

Create a full filesystem backup

For critical debugging situations, create a complete backup of the container’s filesystem:

docker export [container_name_or_id] > container_fs.tar

This creates a tar archive of the entire container filesystem. You can explore it later:

tar -tf container_fs.tar | head -20  # List first 20 files
tar -xf container_fs.tar ./var/log   # Extract just the log directory

Creating a snapshot with docker commit

You can also use docker commit.

This command creates a new Docker image from a stopped container, preserving everything including files, environment, metadata. It’s like taking a snapshot of the container exactly as it stopped.

docker commit [container_name_or_id] debug-snapshot

Now you have a new image called debug-snapshot. You can start it, explore it, even push it to a registry for someone else to examine:

# Start the snapshot container
docker run -it debug-snapshot /bin/bash

# List files inside (from outside)
docker run --rm debug-snapshot ls -la /var/log

# Push to Docker Hub for team analysis
docker tag debug-snapshot yourusername/debug-snapshot
docker push yourusername/debug-snapshot

This works well in a production environment because it preserves the container state completely. You can analyze it later, even if the original container gets removed or rebuilt.

But keep in mind that committed images can be large as they include all container layers. Use them judiciously, and clean up when you’re done with the following command:

docker rmi debug-snapshot

Make future debugging easier

For smoother debugging next time, use persistent volumes for logs and data, set up the right logging driver, and consider scripting the preservation steps you use most often.

When a container stops, save the evidence first, then restart it.

The post How to Debug a Stopped Docker Container appeared first on Hongkiat.

This month’s picks cover minimalist CSS frameworks, JavaScript libraries, WordPress tooling, and a handful of AI-focused tools for building faster workflows.

Comark

Comark is a fast Markdown parser built for streamed content. AI-generated text or progressively loaded content renders cleanly as it arrives. It also auto-closes syntax on the fly and supports plugins such as math and code highlighting. It looks especially useful for docs and interactive blogs.

Loftlyy

Loftlyy is a growing database of real-world brand identities. You can browse logos, colors, and design systems from actual companies. It’s a solid reference if you need inspiration without relying on made-up mock brands.

Oat

Oat is a lightweight HTML and CSS UI component library with zero dependencies. It weighs about 8KB in total and uses semantic HTML tags instead of CSS classes. If you’re tired of JS-heavy UI stacks, this goes in the opposite direction.

PicoCSS

PicoCSS is a minimalist framework that styles HTML directly without extra classes. It uses plain CSS, no JavaScript, and adapts automatically to light or dark mode. Good fit if you want a clean system without carrying extra UI baggage.

Agent Skills by Addy Osmani

Agent Skills by Addy Osmani is a collection of production-grade workflows for AI coding agents. It uses slash commands like /spec, /plan, and /ship to enforce specs, tests, code reviews, and security checks. The appeal here is structure, especially for teams that want agents to follow a stricter engineering process.

It works with Claude Code, Cursor, Gemini CLI, and other agents.

Boneyard

Boneyard automatically generates skeleton screens by snapshotting your real UI. You wrap a component in <Skeleton>, run the CLI once, and it captures pixel-perfect rectangles that mirror your actual layout. No manual measurement or hand-tuned placeholders.

If you build React apps and want to reduce layout shift during loading states, this looks like a neat solution, especially at around 7.5KB.

Locker

Locker is an open-source, self-hosted file platform. You can sync files across multiple storage backends, including S3, R2, and local disk. It also searches inside images and PDFs. Looks useful if you want tighter control over file storage without being locked to one vendor.

WordPress Build Tooling

@wordpress/build is a new build tool for WordPress plugins. It replaces Webpack and Babel with esbuild. So multiple scripts, modules, and admin pages build in seconds with almost no configuration.

Gutenberg already uses it, though it is not ready for every plugin yet. Still, it looks worth a look if you follow WordPress tooling closely.

EmDash

EmDash is a new CMS from Cloudflare, built with TypeScript and Astro. It positions itself as an alternative blogging platform to WordPress.

One interesting part is that it runs plugins in sandboxed Cloudflare Workers, while content uses Portable Text JSON instead of HTML or Gutenberg. That alone makes it stand out from the usual CMS stack.

Open Screen

OpenScreen is a free, open-source macOS app for product demos. It records your screen with zoom effects, annotations, and styled backgrounds. No account required. Handy if you want cleaner product recordings without paying for another screen recording app.

Heerich.js

Heerich.js is a minimalist 3D voxel engine that renders to SVG. It builds compositions using boxes, spheres, and lines. Nice pick for generative art on the web or even pen plotter experiments.

WP Studio

wp-studio is a CLI for local WordPress development that lets you create sites, run WP-CLI commands, and publish temporary previews to wp.build. It requires Node.js 22+. Useful if you want a more scriptable local WordPress setup.

pgmicro

pgmicro is an embeddable, single-file database that speaks PostgreSQL but stores data as SQLite. It parses PostgreSQL and compiles it directly to SQLite bytecode. You can run it in memory, on disk, or as a psql server. Interesting option for ephemeral AI workloads or lightweight local PostgreSQL setups.

Testo

Testo is a modern PHP testing framework for PHP 8.2+. It uses attributes instead of magic conventions, supports functions or classes without inheritance, and includes async tools, benchmarks, and a PhpStorm plugin. The pipe-style assertions are a nice touch if you prefer more explicit, type-safe tests.

PHPantom LSP

PHPantom is a fast PHP language server written in Rust. It supports generics, Laravel Eloquent, PHPStan annotations, and conditional return types. It starts in under 1 second with about 59MB of RAM and skips the usual indexing phase. Strong option if you want faster autocomplete without the usual heavyweight setup.

Graffiti

Graffiti is a minimal CSS toolkit with utilities, elements, blocks, and templates. It’s configurable, themeable, and uses zero JavaScript. It also works with any framework or plain HTML. Good candidate if you want a lightweight drop-in CSS layer without dragging in more JS.

Expect

Expect is a testing skill for AI coding agents. It reads your git changes, generates a test plan, and runs it in a real browser with Playwright. It checks for performance issues, security vulnerabilities, broken links, and design regressions, without forcing you to maintain scripts or selectors. It can also run locally or in CI.

agentOS

agentOS is a portable OS for AI agents. Powered by WebAssembly and V8 isolates, it claims ~6ms cold starts at up to 32x lower cost than sandboxes. It also includes filesystem mounting, host tools, and granular security controls. That makes it a notable option for embedding agents directly into backend systems.

Impeccable

Impeccable is a design skill pack for AI coding agents. It teaches design principles across typography, color, layout, and motion. It also includes a CLI and browser extension that detect more than 25 anti-patterns, including gradient text, overused fonts, and nested cards.

It works with Cursor, Claude Code, Gemini CLI, and more. Useful if you want AI-generated UIs with a bit more visual discipline.

Charcuterie

Charcuterie is a visual explorer for Unicode. It lets you browse characters, discover related glyphs, and learn about scripts and symbols. Rendered glyphs are compared in vector space to power visual similarity.

It is still under active development, but it already looks like a more interesting way to explore Unicode than staring at code charts.

The post Fresh Resources for Web Designers and Developers (April 2026) appeared first on Hongkiat.

I tried dozens of approaches before finding what actually worked. Once I got it running, however, the benefits were clear: unlimited usage, zero API fees, and complete data privacy.

Today, you no longer need a supercomputer to run AI models. You don’t need the latest GPU either. What you need is the right model for your hardware and the know-how to run it efficiently.

In this guide, I’ll show you how to do the same.

Know your hardware

Before you download any model, you need to know what your computer can handle. The common mistake I’ve seen is people trying to run a model that exceeds their physical memory. This could trigger “disk swapping” which could make your laptop unresponsive due to the heavy process.

So first, check your system specs:

• VRAM: If you have a dedicated NVIDIA or AMD GPU, check its Video RAM. This is where the model runs for near-instant responses. 8GB VRAM is a solid baseline for hobby use.
• RAM: 16GB is the absolute minimum I’d suggest for a smooth experience. This handles the “offload”. If a model is 10GB and you only have 8GB of VRAM, the remaining 2GB sits here.
• CPU: Modern processors like Intel i5/i7 or Ryzen 5/7 can run smaller models reasonably well, especially with 4-bit quantization.
• Storage: Ensure you have at least 50GB of SSD space. If your internal storage is tight, you can also run LLMs from an external drive with Ollama. Running models off an old-school HDD will result in painful load times.

Pro Tip: Always subtract ~2GB from your total VRAM/RAM to account for your operating system and open browser tabs. If you have 8GB total, you really need to have 6GB for the AI.

Know your needs

With thousands of models available, don’t just chase the highest benchmark scores. If your hardware is limited, focus on models optimized for your specific tasks.

Since we assume that hardware is constrained, I think there are two use cases that you can realistically run on your laptop: text generation and code generation.

• Coding: Specialized models like Qwen2.5-Coder or DeepSeek-Coder are tuned for syntax and logic.
• Creative Writing: Gemma 4 or Mistral variants tend to have a more natural, less “robotic” prose style.

Consider model size vs. quality

The “B” in 3B or 7B stands for Billions of parameters. More parameters usually mean better reasoning, but higher memory costs.

• 1B – 3B models: Extremely fast, low memory, best for basic grammar and simple summaries.
• 7B – 14B models: A practical range for most users. Good reasoning, and they fit in many modern GPUs.
• 30B+ models: Professional-grade reasoning, but requires high-end hardware (24GB+ VRAM).

Quantization helps here. It compresses the model so it fits on consumer hardware with little loss in output quality.

• 4-bit (Q4_K_M): The industry standard. Reduces memory usage by ~70%.
• GGUF: The most user-friendly format. It allows the model to run on both your CPU and GPU simultaneously.

Can MacBook Air M2 with 8GB RAM run LLMs?

Let’s walk through a concrete example.

Say you have a MacBook Air with an M2 chip (8-core CPU) and 8GB of unified memory. You want to use it for text editing, grammar fixing, and light writing assistance.

With 8GB total RAM, you need to reserve about 2GB for macOS and your other applications. That leaves ~6GB for the model. Apple Silicon’s unified memory architecture also helps because the GPU can access the same memory pool.

Based on these constraints and your needs for text editing and grammar tasks, you don’t need an advanced model with high reasoning capabilities. A model with ~3B parameters is more than enough.

So here are your best options:

• Phi-3.5 Mini 3.8B (Q4_K_M): ~2GB RAM, 20-30 tokens/second. A compact model that handles grammar and editing tasks well enough for daily use.
• Llama 3.2 3B Instruct (Q4_K_M): ~2GB RAM, 15-25 tokens/second. Specifically trained for instruction following, great for “fix this sentence” or “rewrite this paragraph” requests.
• Qwen2.5 3B Instruct (Q4_K_M): ~2GB RAM, similar speed. Good multilingual support if you work with multiple languages.

I’d avoid running 7B models on this hardware. They’ll work but will be slower and might cause swapping if you have other apps open.

Using llmfit to find the perfect model

Manual calculations are a good start, but they still involve some guesswork. If you want a clearer read on what your computer can handle, use llmfit. It scans your hardware and shows which models suit your setup. I also covered how llmfit helps you pick the right local LLM for your machine if you want a closer look at what it does.

You can install llmfit with:

# macOS/Linux with Homebrew
brew install llmfit

# Or quick install
curl -fsSL https://llmfit.axjns.dev/install.sh | sh

Then run it to get recommendations:

llmfit

The tool detects your RAM, CPU cores, and GPU VRAM, then scores hundreds of models based on quality, speed, and how well they fit your hardware.

Each recommendation also includes estimated tokens per second, memory usage, and context length, as we can see below.

You can filter and sort by different criteria, which saves hours of manual testing and helps avoid the frustration of downloading models that won’t run on your hardware.

llmfit integrates with your favorite tools

llmfit also works with tools like Ollama and LM Studio, so the recommendations are easier to act on.

Ollama integration

If you’re using Ollama, llmfit can help you narrow down good model options for your setup. If you prefer a desktop UI instead, LM Studio is another good way to run LLMs locally.

For example, if llmfit recommends google/gemma-2-2b-it, you can immediately hit d and it will show you “Ollama” as an option, as seen below:

Once you’ve selected it, it will download the model for Ollama.

llmfit also supports:

• LM Studio
• llama.cpp
• MLX
• Docker Model Runner

What’s next?

Give it a try. Download a small model, run it locally, and see what you can build with your own private AI assistant.

I recommend llmfit if you want to compare options faster. It would have saved me weeks of trial and error when I was starting out.

The first time you get a response from a model running entirely on your computer, you’ll understand why I made the switch.

The post Choosing the Right LLM Models for Your Everyday Laptop appeared first on Hongkiat.

Dinky is a free app, and that already makes it more appealing if you have been comparing it with paid compression tools like Optimage or Permute.

It is a small macOS utility from Derek Castelli that compresses images, videos, and PDFs in one place. You drop files in, it gives you smaller ones back, but the real appeal is how much repetitive file prep it can absorb once you start using presets, watch folders, and automation.

What Dinky Does

Dinky goes beyond the usual image-only utility. It works with JPG, PNG, WebP, AVIF, TIFF, and BMP, then exports images to WebP, AVIF, or lossless PNG. It also compresses videos to MP4 with H.264 or HEVC presets, and shrinks PDFs while either keeping selectable text and links or flattening pages for more aggressive reduction.

• compress images, videos, and PDFs from one app
• drag and drop files, paste from the clipboard, or use a direct URL
• convert images to WebP, AVIF, or lossless PNG, which pairs nicely with this WebP guide
• resize images or target a specific file size
• save presets, watch folders, and use Apple Shortcuts for repeat jobs
• keep originals, move them to backup, or send them to the trash after compression

That makes it a stronger fit for recurring file prep, especially if you would otherwise piece together the same workflow with Quick Actions or guides like this one on batch compressing files on Mac.

Dinky is listed at around 28 MB installed. Built in Swift and SwiftUI, with no Electron and no web views, it still feels like a lightweight native app rather than a wrapped website pretending to be one.

Who This Is For

Dinky makes the most sense for people who compress files constantly and are tired of doing it in a browser tab.

That includes:

• bloggers and publishers preparing web images
• designers exporting client assets
• people dealing with video uploads
• anyone cleaning up PDFs before sending or archiving them
• Mac users who want a native utility instead of an online service

If your workflow already includes repetitive file prep, Dinky looks like one of those tools that could quietly save time every week.

Before You Install It

There are a few practical limits to keep in mind.

First, it is built for macOS 15 Sequoia and later, not as a cross-platform tool.

Second, it is not notarized, so macOS will probably block it the first time you open it. You can still run it, but you will need to approve it through macOS security settings or remove the quarantine flag manually.

Still, if you are comfortable installing indie Mac software, that probably will not scare you off.

The post Dinky Is a Free Mac App That Compresses Images, Videos, and PDFs appeared first on Hongkiat.

But as I looked at the GitHub code, read Cloudflare’s announcement, and saw Matt Mullenweg’s response, I realized this isn’t just another CMS trying to beat WordPress. I think it’s an interesting test of what a CMS could be in a world with AI tools, serverless hosting, and more security worries.

The real question isn’t if EmDash will replace WordPress. It’s whether it shows us what CMS software might look like next.

What is EmDash CMS?

EmDash is Cloudflare’s attempt to make a next version of WordPress. It is written in TypeScript and built on Astro, which runs well on serverless platforms, especially Cloudflare’s own services. It tries to keep WordPress’s flexibility and admin panel, but adds type safety and plugin sandboxing to fix security issues that WordPress has been facing for years.

Prerequisites

Before you start playing with EmDash, here’s what you’ll need:

• Node.js 22+: For local development
• A Cloudflare account: For the best experience (though it can run elsewhere)
• Basic TypeScript knowledge: The entire codebase is TypeScript
• Familiarity with Astro: EmDash uses Astro for frontend rendering

Installation

You have a few options to install EmDash, depending on how you want to try it:

Create a new EmDash site locally

For local development and testing, use the npm command:

npm create emdash@latest

Deploy directly to Cloudflare

If you want to skip local setup and go straight to deployment, use the Cloudflare deploy link:

https://deploy.workers.cloudflare.com/?url=https://github.com/emdash-cms/templates/tree/main/blog-cloudflare

Try the online playground

For a quick look without any setup, check out the EmDash playground.

Plugin security

Here’s where EmDash goes a different way from WordPress.

WordPress plugins can do anything. They can read and write to your database, change files, connect to the internet. This isn’t a mistake. It’s a choice that allows great flexibility.

But this flexibility also has security risks. So EmDash uses a different method called sandboxing where each plugin runs in its own separate space with clear rules about what it can do.

Here’s what a simple email notification plugin looks like:

import { definePlugin } from "emdash";

export default () => definePlugin({
  id: "notify-on-publish",
  version: "1.0.0",
  capabilities: ["read:content", "email:send"],
  hooks: {
    "content:afterSave": async (event, ctx) => {
      if (event.collection !== "posts" || event.content.status !== "published") return;
      
      await ctx.email!.send({
        to: "editor@example.com",
        subject: `New post published: ${event.content.title}`,
        text: `"${event.content.title}" is now live.`,
      });
      
      ctx.log.info(`Notified editors about ${event.content.id}`);
    },
  },
});

See those capabilities?

The plugin can only do what it says it will do. No secret internet connections, no database access beyond what’s allowed. This is a big change from WordPress’s “trust me with everything” approach.

AI-native from day one

While WordPress 7.0 is starting to introduce AI features into the core through its Connectors API, which enables plugins to connect to external AI services, EmDash takes a more direct approach by giving AI agents the ability to actually run your site.

Every EmDash site comes with three core tools that make this possible:

• Agent skills provide clear instructions that guide AI in building plugins, themes, and making changes to your site—like recipes that agents can follow step by step.
• EmDash CLI gives you powerful command-line tools for managing content, uploading files, and creating content types, making automation and scripting much easier.
• Built-in MCP server. This allows AI tools to connect directly to your site, read and update content, and manage everything through a standard protocol.

WordPress is also moving in a similar direction with its official MCP adapter, which connects the Abilities API to the MCP.

The key difference is that in WordPress, this is an additional plugin you install. In EmDash, it’s built in from the start.

Built-in x402 support for monetization

Here’s something WordPress doesn’t have: every EmDash site has x402 support built in. x402 is an open standard for internet payments that lets you charge for content each time someone uses it.

When an AI tool tries to read your content, it gets a “Payment Required” message, pays immediately, and gets access. No subscriptions, no extra coding.

In a time when AI tools are collecting web content, this feels less like an extra feature and more like something you need to survive.

WordPress import support

EmDash wants to make it as easy as possible to migrate your WordPress blogs. So they provide:

1. A WordPress import tool that makes it easy to migrate your site to EmDash, including content and media
2. Tools to convert WordPress custom post types to EmDash collections
3. Agent skills for porting WordPress themes to EmDash

The migration tools work surprisingly well. I imported a test WordPress site in under 10 minutes.

Migration results, however, may vary depending on your site. If you rely on plugins that add custom post types or custom database tables, you may need to migrate them to EmDash manually.

Where EmDash stumbles

No CMS is perfect, and EmDash is no exception. Matt Mullenweg’s feedback, I think, makes some good points:

1. The sandboxing problem

As Matt says, their sandboxing breaks down as soon as you look at what most WordPress plugins do. Complex plugins that work with many systems, handle files, or connect to other services would need so many permissions that the sandbox doesn’t help much.

2. The UI feels strange

It looks a bit like WordPress but not exactly, and some things don’t work right.

3. Vendor lock-in worries

While it can run elsewhere, it works best on Cloudflare.

4. Missing WordPress’s spirit

WordPress runs everywhere, from Raspberry Pis to $0.99/month Indonesian hosting. EmDash? Not really.

5. No community (yet)

WordPress isn’t just software; it’s meetups, WordCamps, tattoos. EmDash has GitHub stars.

And there’s the April 1st announcement date. Cloudflare has a tradition of releasing real products on April Fools’ Day (remember 1.1.1.1?), but it still feels…weird.

Performance and architecture

Inside, EmDash is built on modern technologies:

1. Astro

For showing web pages, a fast framework for content sites.

2. Cloudflare Workers

Serverless code that starts quickly and runs at the edge.

3. D1 Database

Cloudflare’s SQLite-based database for serverless applications.

4. R2 Storage

For image and file storage with no egress fees.

5. Portable Text

Built on top of PortableText, it uses structured JSON content instead of HTML in the database.

The performance is good. Pages load quickly, and the “scale-to-zero” setup means you only pay for what you use.

Should you use EmDash?

Before deciding whether you should use EmDash, here’s a comparison table to help you see the differences between these two systems:

My opinion:

Use EmDash if:

• You’re already using Cloudflare services
• Plugin security worries you a lot
• You want AI tools built in right now
• You’re starting a new site, not moving an old one

Stick with WordPress if:

• You need to run on cheap, shared hosting
• Your site depends on specific WordPress plugins
• Community and available tools matter more than technical details
• You’re not ready to try a v0.1.0 preview version

The Verdict

EmDash is ambitious, well-built, and working on real problems. The sandboxed plugin model is a careful way to handle security, even if it may have limits for complex plugins. The AI tools feel like they’re thinking ahead. The x402 payment system is a smart answer to AI tools collecting web content.

It’s also v0.1.0, announced on April Fools’ Day, and made to work best with Cloudflare. The UI needs work. It doesn’t have WordPress’s “run anywhere” idea that has let millions of people publish online.

Would I use EmDash for my personal blog today? Probably not. WordPress has too many tools and too much community support to ignore.

Would I try it for a new project where I’m already using Cloudflare and want built-in AI tools? Yes, maybe. But I prefer to wait for it to mature a bit more.

EmDash might not be the next WordPress. And that’s fine. It shows an idea of what CMS software might look like in a future with serverless hosting and AI tools. The sandboxing method isn’t perfect (no security method is), but it’s a useful test of how to balance flexibility with safety.

The best part, I think, is that both WordPress and EmDash can learn from each other. WordPress could use better security methods and AI tools. EmDash could learn from WordPress’s focus on making publishing easy for everyone.

Maybe we’re not looking at a replacement, but at two different ways to solve the same basic problem: helping people put content on the web.

The post A look into EmDash CMS appeared first on Hongkiat.

This is not an official Nous Research desktop app.

Hermes Desktop is a separate open-source project created by GitHub user fathah. It sits on top of Hermes Agent and gives it a native interface for setup, chat, and day-to-day management.

It is a third-party desktop companion for Hermes Agent, not the upstream project itself.

What Is Hermes Desktop?

Hermes Desktop is a native app for installing, configuring, and chatting with Hermes Agent without doing everything by hand from the command line.

According to the project’s GitHub repo, it uses the official Hermes install script, stores Hermes under ~/.hermes, and provides screens for chat, sessions, profiles, memory, skills, tools, schedules, and messaging gateways.

That makes it less of a simple wrapper and more of a desktop control panel for Hermes.

What Can the App Do?

Based on the repo and release notes, Hermes Desktop already goes well beyond a basic chat box.

• first-run install for Hermes Agent
• provider and API key setup
• streaming chat with slash commands and tool progress
• token usage and cost tracking
• session search and resume
• separate Hermes profiles
• skill, tool, memory, and saved model management
• scheduled tasks
• messaging gateway setup for multiple platforms
• logs, backups, imports, and debug tools

Some of that will sound familiar if you have followed tools that make it easier to chat with your bot on Telegram instead of keeping everything locked inside a terminal window.

Why This Stands Out

Hermes Agent itself is powerful, but the official experience is still heavily CLI-driven. That works fine for technical users, but it adds friction for anyone who mainly wants to install the agent, choose a provider, open a chat window, and get started.

Hermes Desktop smooths out that process. Instead of wiring everything together manually, you get one interface for setup, chat, profiles, memory, scheduling, and gateway integrations.

That also makes Hermes easier to compare with other local-first setups, especially if you are already familiar with getting started with Ollama and want something that feels more like a complete agent workspace.

What Providers and Platforms Are Supported?

The repo says Hermes Desktop supports a mix of hosted and local model providers, including mainstream API services and OpenAI-compatible local endpoints.

It also ships with local presets for tools like LM Studio, Ollama, vLLM, and llama.cpp.

As for the app itself, current releases are available for macOS and Linux. The repo lists macOS downloads as .dmg, while Linux builds are offered as .AppImage and .deb.

What It Is Not

The branding can blur together fast.

Hermes Agent is the upstream project from Nous Research. Hermes Desktop is a separate interface layer built around it.

• Hermes Agent is the core agent
• Hermes Desktop is a third-party native app for using it

The difference is useful if you want to know what is official, who maintains the app, and where feature requests or bug reports should go.

Should You Use It?

If you are curious about Hermes Agent but the command-line setup feels like unnecessary friction, Hermes Desktop looks like a strong entry point.

It gives Hermes a real interface, keeps the upstream install flow, and exposes a lot of the project’s useful features without forcing you to memorize commands first.

Just do not mistake it for an official Nous Research desktop release. It is a community-built project around Hermes Agent, and that is part of why people are paying attention to it.

The post Hermes Desktop Is a GUI for Hermes Agent appeared first on Hongkiat.

That is where a lot of AI confusion starts. People hear terms like LLM, 31B, 4-bit, GGUF, LoRA, or uncensored model, and the whole thing starts sounding like a niche obsession for people with too many graphics cards.

It is simpler than it sounds. If you can explain Spotify playlists, ZIP files, and the difference between a general doctor and a specialist, you can explain modern AI too.

This is a plain-English way to do that, especially for large language models, local AI, and the weird filenames that make the whole space look harder than it is.

Start With This: What Is an LLM?

An LLM, short for large language model, is software trained on a huge amount of text so it can predict what words should come next.

That sounds underwhelming, until you realize human conversation works a lot like that too. We read, listen, absorb patterns, then respond based on what we have seen before.

An LLM does something similar at a much larger scale. It is trained on huge amounts of text, often including books, websites, documentation, code, and other language data. It does not think like a person, but it gets very good at recognizing patterns in language. That is why it can answer questions, summarize documents, write emails, explain code, or help brainstorm ideas.

The easiest way to explain it to a friend is this:

An LLM is like a prediction engine for language. It has read an absurd amount of text and learned how words, ideas, and instructions tend to fit together.

If you want an analogy, picture a friend who has read half the internet and can reply instantly in full sentences. That is the big picture.

To make that more concrete, Gemma is one example of an open model family in the LLM world. It is Google’s base model line that people can download, run, and build on.

Then you get custom versions built on top of a model family like that. SuperGemma is a good example. It usually means someone took Gemma, fine-tuned it, changed its behavior, and often packaged it for easier local use.

So the relationship is simple:

• LLM is the broad category
• Gemma is one model family inside that category
• SuperGemma is a customized version built from that family

That framing helps because a lot of AI terms are not separate inventions. They are often layers. First the model type, then the model family, then the customized version.

Parameters Are the Size of the Brain

When people talk about a model being 7B, 12B, 26B, or 31B, they are talking about parameters.

Parameters are the tiny numerical settings inside the model that got adjusted during training. You can think of them as the model’s internal wiring.

More parameters usually means the model can capture more nuance, hold more patterns, and perform better on harder tasks.

A simple way to explain it:

• a smaller model is like a smart pocket notebook
• a larger model is like a full reference library

Both can be useful. The bigger one usually knows more and handles trickier prompts better, but it also needs more memory and more power to run.

So if someone says they are running a 31B model, they mean it is a fairly large one with about 31 billion parameters.

Dense Models vs MoE Models

A dense model uses all of its brain for every reply. Every time it generates text, all of its parameters are involved.

A Mixture-of-Experts model, usually shortened to MoE, is more selective. It has different specialist parts, and only some of them wake up for a given task.

The restaurant analogy works well here. A dense model is like one chef cooking every dish in the restaurant.

An MoE model is like a kitchen with specialists. If you order pasta, the pasta chef gets involved. If you order sushi, someone else steps in. Not every chef needs to touch every plate, which is why MoE models can feel more efficient. They may have a large total size, but only part of that capacity is active at a time.

If you see something like A4B, it usually means around 4 billion parameters are active for each step, even if the overall model is much larger.

Fine-Tuning Is Giving the Model Extra Lessons

A base model is the general version. It knows a broad range of things, but it is not always great at a specific style or task.

Fine-tuning is what happens when someone takes that base model and trains it further on a narrower set of examples.

That is how a general model becomes better at coding, roleplay, customer support, medical note formatting, or following instructions in a cleaner way.

Think of it like this: the base model is a smart student with a broad education, and fine-tuning is sending that student to extra classes. Maybe they become better at coding, more conversational, or less likely to refuse edgy prompts. The original brain is still there, just shaped in a more specific direction.

LoRA and QLoRA Are the Cheap Way to Customize AI

Training a model from scratch is expensive. That is why most hobbyists and small teams do not build a brand-new model. They start with an existing one and adapt it.

LoRA is one popular way to do that. Instead of retraining the whole model, LoRA keeps most of the original model frozen and adds a much smaller set of trainable layers on top. That cuts the cost dramatically.

A clean analogy: full retraining is rewriting an entire textbook, while LoRA is adding a slim companion booklet that updates or extends the original.

QLoRA goes one step further. It first shrinks the model using quantization, then applies the LoRA-style training on top of that smaller version.

That is a big reason local AI got more accessible. It let regular people fine-tune strong models on hardware that would have been laughably underpowered a few years ago.

Quantization Is Model Compression

Quantization means storing the model’s numbers with lower precision so the file becomes smaller and easier to run.

The plain-English version is that you are compressing the model. Not in the exact same way as a ZIP file, but close enough for a beginner explanation.

A 16-bit model keeps more precision. A 4-bit model uses fewer bits per value, which makes it much smaller and lighter.

You lose some quality, but often not as much as people expect.

That tradeoff is why 4-bit models are so popular. They hit a sweet spot: smaller, faster, and much more practical on laptops.

If you need an analogy, compare it to converting a giant RAW photo into a high-quality JPEG. The file gets much smaller. Some detail is lost, but for everyday use it is often still excellent.

GGUF Is the Ready-to-Run File

Once people start downloading local models, they run into file formats.

GGUF is one of the big ones.

The easiest way to explain it is this: GGUF is a packaging format for local models, especially quantized ones.

It packages the model in a way that tools like llama.cpp and LM Studio can load easily.

For non-technical friends, I would just say:

GGUF is the version of the model that has been packed for convenient local use.

If the full original model is a warehouse full of parts, GGUF is the neatly packed version that is easier to move and run.

How to Read Weird Model Names Without Panicking

This is the part that makes local AI look more mysterious than it really is.

Take a name like:

google/gemma-4-26B-A4B-it

Or:

Jiunsong/supergemma4-26b-uncensored-gguf-v2-Q4_K_M.gguf

It looks chaotic, but it is mostly labels stacked together. Here is how to read them.

1. The Creator Name

The part before the slash tells you who published it.

google means the official release came from Google, while Jiunsong means it is a community release from that user or team.

2. The Model Family

gemma-4 or supergemma4 tells you which model line it belongs to.

That is similar to saying iPhone 16, Galaxy S26, or ThinkPad X1. You can also think of it like a car name such as Honda Civic or BMW 3 Series. It tells you the family and generation before you get into the engine, trim, or extras.

3. The Size

26B means about 26 billion parameters.

That gives you a rough sense of the model’s scale.

4. The Architecture Detail

A4B usually points to the active parameter count in an MoE setup.

So while the full model may be larger, around 4 billion parameters are actively doing work at a time.

5. The Behavior or Tuning

it usually means instruction-tuned. In other words, it was trained to follow prompts and behave more like a helpful assistant.

instruct usually signals the same idea.

uncensored usually means the model has fewer refusal rules.

6. The Format

gguf means it is packaged for local running.

7. The Quantization

Q4_K_M is the quantization method.

For beginners, the key detail is simple: it is a 4-bit version, and that usually means a good balance between quality and file size.

So when someone says they are running SuperGemma 26B Q4_K_M GGUF, what they usually mean is:

I am using a 26-billion-parameter custom Gemma model that has been compressed into a practical local file.

That sentence alone clears up a lot.

What Actually Happens From Lab to Laptop

If you want the full journey in one pass, it usually looks like this:

1. A big lab trains the original base model.
2. They release it publicly, or at least release weights people can use.
3. Other developers fine-tune it, compress it, and package it.
4. You download the version that fits your hardware.
5. A local app such as Ollama or LM Studio runs it on your machine. If you want a more containerized setup, this Docker LLM setup guide shows another route.

That is the pipeline. The AI assistant on somebody’s laptop is often just the final step of a long chain of training, adaptation, and compression.

Why Local AI Clicks for Regular People

For most people, the appeal of local AI comes down to three things.

First, privacy. Your prompts and files can stay on your own machine.

Second, cost. Once the model is downloaded, you are not paying per message.

Third, control. You can choose a model that fits your style, hardware, and tolerance for safety filters.

That is why local AI keeps pulling in curious tinkerers, developers, and people who simply do not want all their work flowing through somebody else’s cloud. For a practical example, running LLMs locally with LM Studio shows why that tradeoff feels worth it for many people.

A Simple Way to Explain It to a Friend

If your friend zones out the moment you say “transformer architecture,” skip the jargon and use this version instead:

AI chatbots like ChatGPT run on language models. These models are trained on huge amounts of text so they can predict and generate useful replies. Bigger models are usually smarter, smaller ones are easier to run, and people often compress or customize them so they can work on regular laptops.

That gets you most of the way there.

If they are still curious, add this:

File names that look scary are usually just labels telling you who made the model, how big it is, whether it was customized, and whether it has been compressed for local use.

That is usually the moment the whole thing stops looking mysterious.

Where Beginners Should Start

If someone wants to try local AI without turning it into a weekend project, I would keep it simple.

Start with Ollama or LM Studio. If you need a practical walkthrough, this guide to running an LLM locally with LM Studio is a useful companion. Then pick an instruction-tuned model. If you are on a decent laptop, a 4-bit quantized model is usually the safest starting point.

If the file name still looks intimidating, break it into parts instead of trying to decode it all at once. That is how most people learn it, one label at a time.

Final Thought

You do not need to understand every acronym in AI to talk about it intelligently. You just need a clean mental model.

An LLM is a language prediction engine trained on a huge amount of text. Parameters tell you how big it is, fine-tuning changes its behavior, quantization shrinks it, and GGUF packages it. Those long model names are mostly just specs.

Once you see it that way, AI gets a lot less mysterious. It starts looking like what it really is: software, packaging, tradeoffs, and a lot of labels.

The post How to Explain AI to a Friend Who Doesn’t Follow Tech appeared first on Hongkiat.