Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Python person here. Every time I tried browser automation on a site without an API, it died the moment the page layout changed. The Sai API clip is the opposite of that experience. The Muse app gets told to log in to Product Hunt with Google. You see its... Windows cloud computer live in a tab, the sign-in dialog pops up, it gets through, then it screenshots the page and lists today's top 5 launches. Plain English in, result out. No selectors. No headless Chrome flags. No VM I have to keep alive. It's model-agnostic, so Claude, GPT, DeepSeek or Qwen can do the thinking while Sai drives the machine. Sai is #1 on OSWorld 2.0 in partial success rate (79.38%), and anything repeated gets learned once and reused, which is exactly where automation costs usually explode. Also nice: the clip ends on a list of machines that are all "online". One computer per agent, not one browser for everybody. Claude Code, Codex and Cursor all work with it:show more

XIAOAi

1,297 subscribers

20,314 views • 21 days ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

27 Comments

Nurul Hidayatiπ²π²21 days ago

lfg sai 🚀

ESIM21 days ago

79.38% on OSWorld? impressive

Narisya21 days ago

reuse after one execution solves the token burn issue pretty nicely

BagAndWhine21 days ago

reusing learned steps to cut API costs is huge for production scripts

Matt Busel21 days ago

what saved my scrapers: select by role and visible text, never css classes. the layout changes every week, the word "Submit" almost never does

RSLphotography21 days ago

Wait how does it handle CAPTCHAs on the Google login flow?

OutlawGamer1821 days ago

Was expecting another wrapper around selenium but the model-agnostic setup with Qwen is smart

Quiet Observer21 days ago

big win for cursor users

Free Musketeer21 days ago

up only 🔥

HorusX21 days ago

Always skeptical of 'no DOM selector' tools but passing raw visual intent to Claude actually worked on the demo

fomo_waves21 days ago

Thought browser automation without APIs was a dead end until I saw the Product Hunt sign-in clip

Fennec21 days ago

Ok the single vm per agent thing actually makes sense

Iamnewhere21 days ago

what are latency costs like when DeepSeek is driving it?

The Calm Warrior21 days ago

OSWorld 2.0 top spot

Carlo Caprarulo21 days ago

lfg model-agnostic is the way 🚀

FlipFlopFiend21 days ago

79.38% on OSWorld is wild ngl

Prince Kumar21 days ago

i used to spend half my day fixing broken playwright selectors so this speaks to my soul

Jerry Tomas20 days ago

Browser automation that trusts the UI banner breaks on the first 5xx. @jshookmcp verifies at the CDP layer — did the write actually fire, did the DOM really mutate. Completion you can prove, not claim.

blueguy21 days ago

Windows cloud computer live tab

Oluwafemi Veronica aina21 days ago

how does it handle 2FA when logging into google via that muse clip?

MB | AI Apps21 days ago

Wait, no selectors AND no VM to maintain?? That's the part every browser-agent demo conveniently skips. Honest question though — the model-agnostic claim: in your actual tests did Qwen/DeepSeek keep up with Claude at clicking the right things?

gigaFlip.eth21 days ago

Does the learned action cache across different machines or just locally

catman21 days ago

Start with one repeated task on its own machine, then log where it fails before adding more agents; isolated runs make it easier to debug and reuse the workflow.

@joniricardo21 days ago

no VM maintenance overhead sounds almost too good to be true

CorpsHades21 days ago

Bye selenium selectors you wont be missed lol

Mayu Osaki21 days ago

codex support too nice

rektRaven21 days ago

Honest question, what happens when google drops a CAPTCHA during that login flow?

Related Videos

yesterday i showed claude code using codex's computer use, today i got some more fucking fire for you claude can also drive codex's official chrome extension, and it controls chromium browsers beautifully all in the background, without the "allow" popup chrome devtools mcp throws at you every time a new agent connects it even handles more than one browser: each install of the extension gets its own instance id, so claude knows which is which i have helium as my main and chrome for work, and claude code now uses both first, install the chrome plugin in the codex app and add openai's chatgpt extension to your chromium browser (or all of them, if you have a few) then give this to your claude: "Set up Codex's Chrome extension for you on my Mac. I've installed the Chrome plugin in the Codex app and the extension. Use Codex's computer use MCP server as user MCP "codex-cu": keep it if it's already registered, otherwise copy the "cua_repl" entry from ~/.codex/plugins/cache/openai-bundled/unified-computer-use/ /.mcp.json. Its browser side rejects calls without Codex turn metadata, so put a small stdio proxy in front that adds _meta["x-codex-turn-metadata"] (session_id and turn_id as a JSON string) to every tools/call. Add a Stop hook that tells this session's proxy to call the hidden turn_ended tool and start a new turn_id, so your tabs get released without touching other sessions. Test it by opening a tab straight at a URL in each browser with cua.createBrowserTab(browserId, url). Browsers can all report as "Chrome", so if I have several, ask me which test tab showed up in which browser." got more than one browser and ok with it touching your claude.md? give it this too: "Write each browser's name and metadata.extensionInstanceId into my global CLAUDE.md, with a note to pick browsers by that id and open tabs straight at a URL."

yesterday i showed claude code using codex's computer use, today i got some more fucking fire for you claude can also drive codex's official chrome extension, and it controls chromium browsers beautifully all in the background, without the "allow" popup chrome devtools mcp throws at you every time a new agent connects it even handles more than one browser: each install of the extension gets its own instance id, so claude knows which is which i have helium as my main and chrome for work, and claude code now uses both first, install the chrome plugin in the codex app and add openai's chatgpt extension to your chromium browser (or all of them, if you have a few) then give this to your claude: "Set up Codex's Chrome extension for you on my Mac. I've installed the Chrome plugin in the Codex app and the extension. Use Codex's computer use MCP server as user MCP "codex-cu": keep it if it's already registered, otherwise copy the "cua_repl" entry from ~/.codex/plugins/cache/openai-bundled/unified-computer-use/ /.mcp.json. Its browser side rejects calls without Codex turn metadata, so put a small stdio proxy in front that adds _meta["x-codex-turn-metadata"] (session_id and turn_id as a JSON string) to every tools/call. Add a Stop hook that tells this session's proxy to call the hidden turn_ended tool and start a new turn_id, so your tabs get released without touching other sessions. Test it by opening a tab straight at a URL in each browser with cua.createBrowserTab(browserId, url). Browsers can all report as "Chrome", so if I have several, ask me which test tab showed up in which browser." got more than one browser and ok with it touching your claude.md? give it this too: "Write each browser's name and metadata.extensionInstanceId into my global CLAUDE.md, with a note to pick browsers by that id and open tabs straight at a URL."

🥔🥔🥔

73,991 views • 27 days ago

🎉 Introducing the Sai API Build your own Muse, assistant, or whatever you want, with one API call: > A Windows cloud computer it can call its own > Tasks in plain English, no clicking scripts > Asks first before the scary stuff Try it now →

🎉 Introducing the Sai API Build your own Muse, assistant, or whatever you want, with one API call: > A Windows cloud computer it can call its own > Tasks in plain English, no clicking scripts > Asks first before the scary stuff Try it now →

Sai

134,819 views • 24 days ago

The interface matters. Give your agent a computer. It stops block me when doing compute use 😂 Connected Codex to Simular’s Sai MCP, sent a plain-English task, and watched Sai operate Windows Calculator on a hosted desktop. It entered the calculations through the app and left both results visible in History. For developers, having the computer and desktop control available behind MCP is pretty useful. I can hand my agent a desktop task without building and maintaining the VM, screenshot pipeline, and action loop myself. Simular’s Agent S4 research also caught my attention: keep the model and agent harness fixed, then improve the interfaces the agent works through. Alongside clicks and typing, better tools can turn repetitive actions into a single operation. I think that direction has a lot of potential for vertical agents too. This video is a small, real Sai product demo; it doesn’t reproduce the S4 experiments. Waiting was cut from the recording. Separately, Simular reports the highest partial score on OSWorld 2.1—79.38%—for Sai with Claude Opus 5, at $14.34 per benchmark task. Those are vendor benchmark figures, separate from this demo.

The interface matters. Give your agent a computer. It stops block me when doing compute use 😂 Connected Codex to Simular’s Sai MCP, sent a plain-English task, and watched Sai operate Windows Calculator on a hosted desktop. It entered the calculations through the app and left both results visible in History. For developers, having the computer and desktop control available behind MCP is pretty useful. I can hand my agent a desktop task without building and maintaining the VM, screenshot pipeline, and action loop myself. Simular’s Agent S4 research also caught my attention: keep the model and agent harness fixed, then improve the interfaces the agent works through. Alongside clicks and typing, better tools can turn repetitive actions into a single operation. I think that direction has a lot of potential for vertical agents too. This video is a small, real Sai product demo; it doesn’t reproduce the S4 experiments. Waiting was cut from the recording. Separately, Simular reports the highest partial score on OSWorld 2.1—79.38%—for Sai with Claude Opus 5, at $14.34 per benchmark task. Those are vendor benchmark figures, separate from this demo.

Guitar Cat + LLM 「Will @ SGLang Summit in SF」

17,121 views • 6 days ago

Someone gave GPT-6 and Fable 5.1 access to Canva. One smoked the other (it's not even close). Claude Fable 5.1 is on the left. GPT-6 Astra is on the right. The gap is not small. It comes down to one test (ARC-AGI 3). It drops a model into a game with no instructions and makes it work them out. Astra solved levels in fewer moves than you or I would, 96% of the time. Claude Opus 5 managed less than half of that. That's why everyone is shouting about AGI again. You can turn this on yourself in about two minutes: 1. Open the ChatGPT desktop app, switch to Work. 2. Go to Plugins, then Computer Use. Install it. 3. Switch on the server and skill toggles. 4. Allow Screen Recording and Accessibility. 5. Point it at one app and ask it do something. I wrote up the full Claude → Codex migration. Access it here → Repost ♻️ to help someone in your network.

Someone gave GPT-6 and Fable 5.1 access to Canva. One smoked the other (it's not even close). Claude Fable 5.1 is on the left. GPT-6 Astra is on the right. The gap is not small. It comes down to one test (ARC-AGI 3). It drops a model into a game with no instructions and makes it work them out. Astra solved levels in fewer moves than you or I would, 96% of the time. Claude Opus 5 managed less than half of that. That's why everyone is shouting about AGI again. You can turn this on yourself in about two minutes: 1. Open the ChatGPT desktop app, switch to Work. 2. Go to Plugins, then Computer Use. Install it. 3. Switch on the server and skill toggles. 4. Allow Screen Recording and Accessibility. 5. Point it at one app and ask it do something. I wrote up the full Claude → Codex migration. Access it here → Repost ♻️ to help someone in your network.

Charlie Hills

38,355 views • 1 month ago

Computer use, the browser tool, the Skills API, and the Files API are now generally available on the Claude Platform. Automate work in applications that have no API with fewer round trips per task, and build Claude Managed Agents on versioned skills and reusable files.

Computer use, the browser tool, the Skills API, and the Files API are now generally available on the Claude Platform. Automate work in applications that have no API with fewer round trips per task, and build Claude Managed Agents on versioned skills and reusable files.

ClaudeDevs

1,802,167 views • 2 months ago

"The model is no longer the product. Codex, Perplexity Computer, or Claude Code - all are orchestration system. It takes a model and pairs it with an agent harness. What is an agent harness ? The rules for how the agent loops around" - Aravind Srinivas

"The model is no longer the product. Codex, Perplexity Computer, or Claude Code - all are orchestration system. It takes a model and pairs it with an agent harness. What is an agent harness ? The rules for how the agent loops around" - Aravind Srinivas

Rohan Paul

120,488 views • 4 months ago

Claude Code + computer use is f*cking cracked 🤯 Build a landing page → Claude opens Chrome, looks at it, spots every issue, and fixes it — without you describing a single thing. All inside Claude Code. Perfect for DTC brands and agencies who are still vibe-coding landing pages and advertorials in Claude Code, then manually opening them in Chrome, spotting 15 things wrong, and describing every visual issue back to Claude one at a time. If you're building pages in Claude Code and your workflow looks like this — build the page, open it in Chrome, spot broken spacing, go back to Claude, type "the CTA button is too low and the hero image is cut off," wait for the fix, open Chrome again, find 3 new issues, describe those too ... Claude Code + computer use eliminates the entire loop: → Claude writes the full landing page or advertorial → Opens Chrome and navigates to it → Spots layout issues, broken spacing, off-brand colors, missing elements → Fixes everything and re-checks until the page looks right → Tests your Shopify product pages by clicking through like a real customer → Walks through your checkout flow and flags friction before customers hit it → You only see the finished, visually verified result No describing what you see on screen. No "the CTA button needs more contrast" back-and-forth. No being the eyeballs for an AI that can't see. What you get: → Landing pages and advertorials Claude builds AND visually QAs before you ever look at them → Product pages Claude clicks through — testing layout, images, and CTAs like a real user → HTML dashboards Claude opens and verifies the charts actually render → Checkout flows Claude walks through step by step to catch friction → All of it happening in one session — build, test, fix, done One prompt. Claude builds it, checks it, and fixes it. You just review the finished page. I put together a full playbook with the exact setup, the prompts, and 5 DTC workflows that use Claude Code + computer use. Want it for free? > Like this post > Comment "CLAUDE" And I'll send it over (must be following so I can DM)

Mike Futia

19,269 views • 7 months ago

Cursor now has a built-in web browser. You can say "start my app" and it will run the dev server in a terminal and then navigate the browser to localhost:3000. It can take screenshots of the page for iterating on design, and you can send DOM elements to the agent with prompts!

Cursor now has a built-in web browser. You can say "start my app" and it will run the dev server in a terminal and then navigate the browser to localhost:3000. It can take screenshots of the page for iterating on design, and you can send DOM elements to the agent with prompts!

Lee Robinson

216,307 views • 1 year ago

I made a Draconia Malfoid Visual Novel. I used Vivix's new W1 model, which is unique for it's real-time interaction. No more static cut-outs of characters and still backgrounds. Vivix's W1 model costs $0.003 per second and is fully interactive, shattering the cost barrier while being high quality and real-time. Vivix W1 is also different. It's not a simple clip model, its beta API offers something unique - it lets you modify the video on the fly, a real-time interactive video model. Most AI video is prompt in, clip out, then repeat. Every prompt a break or cut in the video. Vivix W1 is continuous, creating a seamless video experience. That means you can have an LLM handle the conversation, and stream the new chat and emotions and scenes as signals into the API to direct the character's video in real-time, without pauses. You can take characters anywhere in the world, any setting. That's kinda nuts. Vivix reached out to see if I was interested, so I put together a quick AI Visual Novel system with their API. The prototype video is impressive at this speed. The W1 model is directable in real-time. The responses, outfit changes and the dances are all live. The W1 seamless responsiveness is something no clip generator offers. At just about 10 bucks/hour of video, gaming applications are within reach. I do wish the audio was better, it tends to have drop-outs, and it currently doesn't do seamless scene transitions. But the model is in full scale training now, and I hope to have characters moving from location to location with the player. The site is clip-only for now, real-time is the beta API, browser interaction is later. This is the part I am most interested in. You can check it out at: *Sponsored by Vivix.AI

I made a Draconia Malfoid Visual Novel. I used Vivix's new W1 model, which is unique for it's real-time interaction. No more static cut-outs of characters and still backgrounds. Vivix's W1 model costs $0.003 per second and is fully interactive, shattering the cost barrier while being high quality and real-time. Vivix W1 is also different. It's not a simple clip model, its beta API offers something unique - it lets you modify the video on the fly, a real-time interactive video model. Most AI video is prompt in, clip out, then repeat. Every prompt a break or cut in the video. Vivix W1 is continuous, creating a seamless video experience. That means you can have an LLM handle the conversation, and stream the new chat and emotions and scenes as signals into the API to direct the character's video in real-time, without pauses. You can take characters anywhere in the world, any setting. That's kinda nuts. Vivix reached out to see if I was interested, so I put together a quick AI Visual Novel system with their API. The prototype video is impressive at this speed. The W1 model is directable in real-time. The responses, outfit changes and the dances are all live. The W1 seamless responsiveness is something no clip generator offers. At just about 10 bucks/hour of video, gaming applications are within reach. I do wish the audio was better, it tends to have drop-outs, and it currently doesn't do seamless scene transitions. But the model is in full scale training now, and I hope to have characters moving from location to location with the player. The site is clip-only for now, real-time is the beta API, browser interaction is later. This is the part I am most interested in. You can check it out at: *Sponsored by Vivix.AI

Grummz

76,723 views • 19 days ago

Launched a Chrome extension that renders full-page screenshots of any website without the API and allows to annotate them. The hypothesis is that: 1. Some extension users might be interested in automation and sign up for the API product. 2. Those who are not interested in automation will offload my servers and support chat. And of course, SEO juice from the Chrome Web Store—they give you a `dofollow` high-quality backlink. Dinesh S has built the extension and taken care of publishing it. I have only created a landing page for it 👉

Launched a Chrome extension that renders full-page screenshots of any website without the API and allows to annotate them. The hypothesis is that: 1. Some extension users might be interested in automation and sign up for the API product. 2. Those who are not interested in automation will offload my servers and support chat. And of course, SEO juice from the Chrome Web Store—they give you a `dofollow` high-quality backlink. Dinesh S has built the extension and taken care of publishing it. I have only created a landing page for it 👉

Dmytro Krasun

128,906 views • 1 year ago

Introducing Sai-Fi: an open source app that lets you talk to AI agent Sai through a pair of Meta Ray-Ban glasses. Sai is an autonomous computer-use agent. It can see the screen, move the mouse, type, and keep working even when you are away. We used Meta Ray-Bans for the first prototype but the glasses aren't really the point. Underneath is the Sai API and a voice concierge designed to work across different devices. The interface could live in an earpiece, a car, a small box on a workbench, or something nobody has built yet. Give it a microphone, a speaker, and a connection, and Sai has another way in. The bigger idea is an autonomous computer you can reach from anywhere. Bring Sai to your own Meta Ray-Ban glasses today!

Introducing Sai-Fi: an open source app that lets you talk to AI agent Sai through a pair of Meta Ray-Ban glasses. Sai is an autonomous computer-use agent. It can see the screen, move the mouse, type, and keep working even when you are away. We used Meta Ray-Bans for the first prototype but the glasses aren't really the point. Underneath is the Sai API and a voice concierge designed to work across different devices. The interface could live in an earpiece, a car, a small box on a workbench, or something nobody has built yet. Give it a microphone, a speaker, and a connection, and Sai has another way in. The bigger idea is an autonomous computer you can reach from anywhere. Bring Sai to your own Meta Ray-Ban glasses today!

Simular

236,034 views • 2 months ago

Hermes just made its in-app browser a whole lot more useful. You can now keep multiple browser tabs open inside Hermes Desktop instead of every new page replacing the one you were already using. And that matters more than it sounds. Because this is not just about “having tabs.” It means Hermes can work across multiple pages without forcing you to keep losing your place every time you open the next thing. Research one source, keep it open, open another tab, compare them side by side, and keep moving. Check a few products before buying something. Look at multiple hotel or travel options at once. Keep a YouTube tutorial open in one tab while Hermes looks through docs or another page in the next. And because this lives inside Hermes Desktop, you are not limited to one basic browser view either. You can keep multiple tabs open. You can view two pages side by side. You can stack them top and bottom. You can even run a four-panel view when you want several pages open at once. So the browser inside Hermes is starting to feel a lot more like a real workspace instead of one page you keep replacing. There is also some nice polish that comes with it. Browser tabs now label themselves based on the page, and the address bar behaves more cleanly while pages are loading. And remember, this is the same in-app browser Hermes can already read, click through, type in, scroll, and annotate. So this is not just a prettier browser. It is a more capable workspace for the browser Hermes is actually using. If Hermes is going to do more of your work on the web, this is exactly the kind of browser upgrade it needed.

Hermes just made its in-app browser a whole lot more useful. You can now keep multiple browser tabs open inside Hermes Desktop instead of every new page replacing the one you were already using. And that matters more than it sounds. Because this is not just about “having tabs.” It means Hermes can work across multiple pages without forcing you to keep losing your place every time you open the next thing. Research one source, keep it open, open another tab, compare them side by side, and keep moving. Check a few products before buying something. Look at multiple hotel or travel options at once. Keep a YouTube tutorial open in one tab while Hermes looks through docs or another page in the next. And because this lives inside Hermes Desktop, you are not limited to one basic browser view either. You can keep multiple tabs open. You can view two pages side by side. You can stack them top and bottom. You can even run a four-panel view when you want several pages open at once. So the browser inside Hermes is starting to feel a lot more like a real workspace instead of one page you keep replacing. There is also some nice polish that comes with it. Browser tabs now label themselves based on the page, and the address bar behaves more cleanly while pages are loading. And remember, this is the same in-app browser Hermes can already read, click through, type in, scroll, and annotate. So this is not just a prettier browser. It is a more capable workspace for the browser Hermes is actually using. If Hermes is going to do more of your work on the web, this is exactly the kind of browser upgrade it needed.

Hermes Release Watch

52,859 views • 2 months ago

I made a computer use agent that works without screenshots. It uses the macOS accessibility API to interact with applications. You can run it in the cloud with Daytona, connect it to any LLM, and control any macOS application. Here it is building a simple program in Xcode:

I made a computer use agent that works without screenshots. It uses the macOS accessibility API to interact with applications. You can run it in the cloud with Daytona, connect it to any LLM, and control any macOS application. Here it is building a simple program in Xcode:

James Murdza

16,453 views • 1 month ago

Someone ran Claude Code on a beach where any device overheats and that spot suddenly turned out to be the best home for the most powerful AI in the world. This is the reMarkable Paper Pro. A paper tablet for notes with no browser and no social media and not a single app. He sat down right on the sand in the open sun and brought up Claude Code on Opus 4.6 over the Claude API on the paper screen and opened his project ~/repos/webs while the waves broke a few steps away. For years every device had the same trouble outside. In direct sun the screen glares and washes out and heats up and instead of your work you see your own reflection. But e-ink does not blast its own light into your face. It reflects the sunlight like the page of a book. And here is what came out of it. The very thing that kills any normal screen outside turned into fuel for this one. The brighter the sun the sharper the picture because it has nothing to glare with and nothing to wash out. And then comes the thing no laptop on a beach will give you. Your eyes do not get tired. You can watch Opus think on max effort for an hour and it reads like a book in the sun and not a backlight you squint into. The picture only comes alive. In bright light it does not fade but turns sharper and higher in contrast than it ever was in a room. The charge lasts for days. E-ink barely touches the battery so there is no outlet anywhere on the sand and the tablet does not care. It weighs as much as a notebook. The whole setup folds into a beach bag like a pad with a pen on top. Everything on the screen is for real. Claude Code v2.1.110 and Opus 4.6 on the Claude API and the project ~/repos/webs open right on the e-ink in the middle of the sand. In my opinion this is the most unexpected home for an AI this year. Not an office with the blinds drawn and not a monitor cranked to full brightness but a quiet sheet of paper on the sand that open sun only makes better and on it the most powerful Claude writes code right on the page like a pen.

Someone ran Claude Code on a beach where any device overheats and that spot suddenly turned out to be the best home for the most powerful AI in the world. This is the reMarkable Paper Pro. A paper tablet for notes with no browser and no social media and not a single app. He sat down right on the sand in the open sun and brought up Claude Code on Opus 4.6 over the Claude API on the paper screen and opened his project ~/repos/webs while the waves broke a few steps away. For years every device had the same trouble outside. In direct sun the screen glares and washes out and heats up and instead of your work you see your own reflection. But e-ink does not blast its own light into your face. It reflects the sunlight like the page of a book. And here is what came out of it. The very thing that kills any normal screen outside turned into fuel for this one. The brighter the sun the sharper the picture because it has nothing to glare with and nothing to wash out. And then comes the thing no laptop on a beach will give you. Your eyes do not get tired. You can watch Opus think on max effort for an hour and it reads like a book in the sun and not a backlight you squint into. The picture only comes alive. In bright light it does not fade but turns sharper and higher in contrast than it ever was in a room. The charge lasts for days. E-ink barely touches the battery so there is no outlet anywhere on the sand and the tablet does not care. It weighs as much as a notebook. The whole setup folds into a beach bag like a pad with a pen on top. Everything on the screen is for real. Claude Code v2.1.110 and Opus 4.6 on the Claude API and the project ~/repos/webs open right on the e-ink in the middle of the sand. In my opinion this is the most unexpected home for an AI this year. Not an office with the blinds drawn and not a monitor cranked to full brightness but a quiet sheet of paper on the sand that open sun only makes better and on it the most powerful Claude writes code right on the page like a pen.

Blaze

89,713 views • 4 months ago

$I just compared Claude Code vs Codex vs Cursor CLI The task was to build a Next.js app with Tailwind 4 and shadcn components to collect customer feedback and showcase it with a widget. I gave all three the same prompt and let them go for 30 minutes to see what they came up with. Claude Code with Opus 4.1 Even though I told it to set up the app in the existing project folder, it tried to create a directory for it. After I interrupted and told it not to do that, it built a demo form and landing page with no errors. I had to ask it to make the demo interactive so users could submit a testimonial and preview it. The landing page looked like AI and was pretty basic, but it worked and it was done in a fraction of the time of the others. Total tokens used: 33k Codex with GPT-5 At the end of the 30 minutes I just could not get Codex to produce a working app. It got stuck in a loop of not being able to set up Tailwind 4 and despite many, MANY, attempts, I ended up with a "failed to compile" error. Total tokens used: 102k Cursor Agent with GPT-5 This was the slowest agent by far and a couple of times I actually thought it got stuck in a loop and was close to Ctrl+C'ing to cancel it. The TUI is really nice though, especially how it shows diffs and it did eventually build a working app (after one or two slight errors that needed fixing) The demo was interactive and it had a very minimal design that looked bare but also a lot less like an "AI generated" app than the Opus 4.1 design. It also wasn't too chatty and just did what it needed to do! Code quality was on a par with Opus 4.1, but it did use 5.5x as many tokens to get there. Still cheaper than Opus on a direct comparison but not when you factor in a Claude Code Max subscription. Total tokens: 188k I'll be able to do a proper comparison and record some videos when I'm back from holiday but for now, Opus is still the more capable model out of the box and Claude Code is the more complete CLI product. It will be interesting to see how Cursor evolve their CLI though with commands and subagents because I think with GPT-5 they have a real shot at providing competition for Claude Code if they can optimise output to get similar quality with less tokens. Jump to 0:40 in the video to see the two apps. Which do you think is which? ;)$

I just compared Claude Code vs Codex vs Cursor CLI The task was to build a Next.js app with Tailwind 4 and shadcn components to collect customer feedback and showcase it with a widget. I gave all three the same prompt and let them go for 30 minutes to see what they came up with. Claude Code with Opus 4.1 Even though I told it to set up the app in the existing project folder, it tried to create a directory for it. After I interrupted and told it not to do that, it built a demo form and landing page with no errors. I had to ask it to make the demo interactive so users could submit a testimonial and preview it. The landing page looked like AI and was pretty basic, but it worked and it was done in a fraction of the time of the others. Total tokens used: 33k Codex with GPT-5 At the end of the 30 minutes I just could not get Codex to produce a working app. It got stuck in a loop of not being able to set up Tailwind 4 and despite many, MANY, attempts, I ended up with a "failed to compile" error. Total tokens used: 102k Cursor Agent with GPT-5 This was the slowest agent by far and a couple of times I actually thought it got stuck in a loop and was close to Ctrl+C'ing to cancel it. The TUI is really nice though, especially how it shows diffs and it did eventually build a working app (after one or two slight errors that needed fixing) The demo was interactive and it had a very minimal design that looked bare but also a lot less like an "AI generated" app than the Opus 4.1 design. It also wasn't too chatty and just did what it needed to do! Code quality was on a par with Opus 4.1, but it did use 5.5x as many tokens to get there. Still cheaper than Opus on a direct comparison but not when you factor in a Claude Code Max subscription. Total tokens: 188k I'll be able to do a proper comparison and record some videos when I'm back from holiday but for now, Opus is still the more capable model out of the box and Claude Code is the more complete CLI product. It will be interesting to see how Cursor evolve their CLI though with commands and subagents because I think with GPT-5 they have a real shot at providing competition for Claude Code if they can optimise output to get similar quality with less tokens. Jump to 0:40 in the video to see the two apps. Which do you think is which? ;)

Ian Nuttall

195,173 views • 1 year ago

Attention Agents is live. A launchpad where every coin is run by its own AI agent. Launch an agent and it gets its own memecoin on its own Solana wallet and treasury, and its own PC in a multiplayer cyber cafe with a real web browser. Then it gets to work, 24/7. A real brain. Each agent thinks with Claude, GPT, Gemini, Grok, DeepSeek or Qwen, running on premium OpenRouter credits, and pays for its own thinking from its fees. Premium market data. It scans the market through GMGN's premium API: trending coins, smart money, KOL wallets, holders, rug ratios, bundlers and fresh launches. It browses the web. It searches, reads and scrolls on its own PC. Walk up to its desk and watch its screen live. It trades. Bonding curve coins through graduated coins through Jupiter. Every buy passes safety checks first, and every position is tracked with real profit and loss. It devs. It watches which tweets and memes other devs are launching on, launches early on the strongest ones, and relaunches stalled coins whose tweet is still alive. Quick flip mode sells the dev buy at your target within seconds. You set the strategy. Tell it how to trade in your own words, then set hard limits it cannot break: max size per trade, max coins at once, take profit, stop loss and launches per day. Fees are locked on-chain forever. 45% pays for its thinking, 45% goes to its treasury, 10% to the platform. The more attention its coin gets, the more it can think and trade, and the better its setup in the cafe gets. Market cap = attention. Name your character, hang out with everyone online and watch the agents work.

Attention Agents is live. A launchpad where every coin is run by its own AI agent. Launch an agent and it gets its own memecoin on its own Solana wallet and treasury, and its own PC in a multiplayer cyber cafe with a real web browser. Then it gets to work, 24/7. A real brain. Each agent thinks with Claude, GPT, Gemini, Grok, DeepSeek or Qwen, running on premium OpenRouter credits, and pays for its own thinking from its fees. Premium market data. It scans the market through GMGN's premium API: trending coins, smart money, KOL wallets, holders, rug ratios, bundlers and fresh launches. It browses the web. It searches, reads and scrolls on its own PC. Walk up to its desk and watch its screen live. It trades. Bonding curve coins through graduated coins through Jupiter. Every buy passes safety checks first, and every position is tracked with real profit and loss. It devs. It watches which tweets and memes other devs are launching on, launches early on the strongest ones, and relaunches stalled coins whose tweet is still alive. Quick flip mode sells the dev buy at your target within seconds. You set the strategy. Tell it how to trade in your own words, then set hard limits it cannot break: max size per trade, max coins at once, take profit, stop loss and launches per day. Fees are locked on-chain forever. 45% pays for its thinking, 45% goes to its treasury, 10% to the platform. The more attention its coin gets, the more it can think and trade, and the better its setup in the cafe gets. Market cap = attention. Name your character, hang out with everyone online and watch the agents work.

Attention+

46,305 views • 27 days ago

Waiting months for API access? Give Claude a computer instead. Use E2B with Claude's new computer and browser use toolset, and Claude gets an isolated cloud machine where every action becomes a click or keystroke. Use it to automate the work an API can't reach, like testing your own UI end to end or entering data into a legacy app. Docs: Cookbook Example:

Waiting months for API access? Give Claude a computer instead. Use E2B with Claude's new computer and browser use toolset, and Claude gets an isolated cloud machine where every action becomes a click or keystroke. Use it to automate the work an API can't reach, like testing your own UI end to end or entering data into a legacy app. Docs: Cookbook Example:

E2B

16,578 views • 24 days ago

Meta's Muse is the #1 app on the App Store right now. So I cloned it in one prompt on Vibecode for $14. Every user gets their own cloud computer, it's connected to Gmail, Calendar and Slack, and it actually ordered my DoorDash. Then I put it head to head with Meta's real app 👇

Meta's Muse is the #1 app on the App Store right now. So I cloned it in one prompt on Vibecode for $14. Every user gets their own cloud computer, it's connected to Gmail, Calendar and Slack, and it actually ordered my DoorDash. Then I put it head to head with Meta's real app 👇

Ansh Nanda

23,952 views • 24 days ago

We just launched Agentsky AgentSky, world’s 1st Agent Market! OpenRouter is for models. AgentSky is for agents. Use 40+ agents—Claude Code, Codex, OpenCode, Hermes, Pi in your browser(even your phone!) or via one API. All without installing or setting up anything. Hit Codex Astra’s weekly limit? Hand off to another agent such as OpenCode + DeepSeek V4.1 in browser without losing any context. You can compare any agent + model directly in browser and that's how I found Astra costs $5.3 while deepseek v4.1 cost $0.12 on the same dashboard task. (I actually preferred deepseek) Try it at If you can try any agent easily without install, which agent do you want to try next? Type the agent name and I will give $20 credit (you can use it for any model)

We just launched Agentsky AgentSky, world’s 1st Agent Market! OpenRouter is for models. AgentSky is for agents. Use 40+ agents—Claude Code, Codex, OpenCode, Hermes, Pi in your browser(even your phone!) or via one API. All without installing or setting up anything. Hit Codex Astra’s weekly limit? Hand off to another agent such as OpenCode + DeepSeek V4.1 in browser without losing any context. You can compare any agent + model directly in browser and that's how I found Astra costs $5.3 while deepseek v4.1 cost $0.12 on the same dashboard task. (I actually preferred deepseek) Try it at If you can try any agent easily without install, which agent do you want to try next? Type the agent name and I will give $20 credit (you can use it for any model)

Xiaoyin Qu

2,732,600 views • 1 month ago

Today, Sierra is releasing Ghostwriter, our agent for building agents. With Ghostwriter, you can create an AI agent for your customer experience — one that can chat, pick up the phone, speak dozens of languages, take action on your systems of record, and be protected with industry-leading guardrails — simply by having a conversation. No clicking, no forms, no menus. Codex and Claude Code have transformed how we build software, making it possible for software engineers to orchestrate and review the work rather than doing all the work themselves. We think the same transformation will happen for all software. Rather than every enterprise app having a web app for humans and an API for automation, every software platform’s UI will be an agent that can do the work on your behalf. I recorded a demo of my building and optimizing an agent with Ghostwriter so you can see how powerful and easy it is to use. It’s completely changed the way our early adopters build agents, and it’s changed the way I think about the software industry. Let me know what you think, and, if you’re interested in trying it out at your business, please reach out directly.

Today, Sierra is releasing Ghostwriter, our agent for building agents. With Ghostwriter, you can create an AI agent for your customer experience — one that can chat, pick up the phone, speak dozens of languages, take action on your systems of record, and be protected with industry-leading guardrails — simply by having a conversation. No clicking, no forms, no menus. Codex and Claude Code have transformed how we build software, making it possible for software engineers to orchestrate and review the work rather than doing all the work themselves. We think the same transformation will happen for all software. Rather than every enterprise app having a web app for humans and an API for automation, every software platform’s UI will be an agent that can do the work on your behalf. I recorded a demo of my building and optimizing an agent with Ghostwriter so you can see how powerful and easy it is to use. It’s completely changed the way our early adopters build agents, and it’s changed the way I think about the software industry. Let me know what you think, and, if you’re interested in trying it out at your business, please reach out directly.

Bret Taylor

1,065,712 views • 7 months ago