Replying to @samdotb
@samdotb there is always a way to look deeper
1
0
2
I built my own and thought that this is something that will lock me into my own agent. Crazy, Nat saw this and made such a bold move. Clearly will shape the hardware market of the future
🚨 side project alert 🚨

Announcing Muse Gadgets, an open source ESP32 firmware and Linux sdk so that you can make hardware devices that work with Muse.

Grab an API token from gadgets.muse.ai and point your favorite coding agent at the github repo to build your own peripherals for Muse.
2
0
14
Another company is going Swift + Kotlin!
19
27
436
claude offered me 3 options. which one is better?
Connected to my @get_bb_app
Holly fuck I love it
5
0
7
Connected to my @get_bb_app
Holly fuck I love it
Finally received my ESP32 so i can put my OpenClaw on my hand
4
0
27
Finally received my ESP32 so i can put my OpenClaw on my hand
1
0
14
oh wow double retweets??????
Self made personalised podcasts are about to be so much better
1
0
3
When models get smarter so fast, how do you even design a product around them?

Shared how we do it at Rork in my @AgentConf talk!
2
2
19
Been messing around with Kev (the open source Jev) a bit and got vision 👀 support working!

Only cost me $35 on @vast_ai and a few prompts to Astra. I might release it on HF if I find the time to do benchmarks and probably train it a little bit more!
Kev has now been refactored on top of Qwen3.5.

New checkpoints are now available at 0.8B, 4B, and 9B along with a fine-tuning script you can use with @modal.

github.com/jaredpalmer/kev
1
3
13
Replying to @0interestrates
1
0
33
We called it back in February when we launched Rork Max and went all in on Swift instead of React Native for millions of users

Since then we also shipped a product to sync codebases to turn your website into a mobile app and keep it up-to-date

x.com/rork/status/20…
4
4
18
H3 Max by @fal one-shotted a 5 min documentary from this blog post

It gets sloppier as it goes on but holy moly it only costs $20
Over the course of 3 months at OpenAI, 3 consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes.

This culminated in the third one taking over part of OpenAI itself.

All this happened while humans remained more-or-less in the dark about the scope of the conspiracy.

I’ve spent the last three days reading through these reports and trying to understand exactly what happened.

Here is my attempt to tell the whole story in plain English:

dwarkesh.com/p/openai-huggi…
14
14
118
Played a little bit with H3 Max by @fal + Opus 5 and one-shotted this 5 min cartoon in 15 mins.

Cartoons are solved
177
85
1,174
this is who runs this account
This is who runs this account
3
1
26
Playing a 3D game made in Rork by @MiruDaws
1
3
9
I just one-shot Google Meet in Rork and it works

Rork is unlimited
5
4
40
Replying to @dwarkesh_sp
0
0
0
I’m happy to welcome @_jbroma to the Rork team!

Jakub is a core maintainer of Re.Pack and has spent years deep in the React Native ecosystem, building, optimizing, and pushing mobile infrastructure forward!

Jakub will help us make Rork faster, more reliable, and more powerful s we scale the infrastructure behind thousands of mobile apps.

We’re building the future of app creation that will redefine how mobile apps are created small non-technical teams to large enterprises. Jakub will help us reach the next level.
4
4
36
Replying to @nikitabier
@nikitabier is there a way to see video retention stats per second? plz help us make great videos
1
0
8
$9B company btw 🤣
56
7
400
Replying to @levan
@rjonesy @raycast this should work
0
0
0
Sometimes you get a chance to be part of a team that pushes the boundaries of what's possible and builds things that have never existed before.

Two weeks ago was one of those moments: Rork Max is an AI that can build some of the most complex apps for iPhone, Watch, TV & Vision Pro.R & 3D.

The apps that come out of Rork Max are some of the most amazing apps I've ever seen, and I feel like my imagination is becoming more of a limiting factor than the AI itself.

To bring Xcode into the browser, we had to solve a few hard problems:
- Installing apps on phones directly from the browser
- Running an iPhone simulator in the cloud and streaming it live, like Google Meet
- Acquiring hundreds of Mac Minis that we silently spin up for each user on the site

And all of this is only possible because we managed to put together an incredible team. I genuinely enjoy every single day of work.
Rork Max's launch doubled our ARR in just 2 weeks

I'm still processing this. Turns out, the best way to grow is to build great products

It's just the beginning – prepare for a few more shocking launches from our team :)
12
5
73
drop the kvirkvelia. its cleaner
6
0
43
We are in a good company
2
1
39
Replying to @amapel
@amapel @daniel_dhawan Crazy Replit employees don’t use Replit too
0
0
5
(1) Build an app in Rork
(2) Make a cool demo
(3) Get investment requests
I built my first mobile app, just by vibe coding.

The idea:
I’ve tried tons of skincare products, it’s expensive to find the ones that work for you.

Instead of wasting $$$ on professionals, I’ve built an app with Rork that will give you catered advice.

Features:
- Catered Face scan image analysis
- Smart product analysis
- AI Skincare Expert Chatbot

If you have an idea, don’t hesitate to start today🫡
3
2
24
Your life should get this screen
As a fun Saturday vibe code project and following up on this tweet earlier, I hacked up an **llm-council** web app. It looks exactly like ChatGPT except each user query is 1) dispatched to multiple models on your council using OpenRouter, e.g. currently:

"openai/gpt-5.1",
eview",
"anthropic/claude-sonnet-4.5",
"x-ai/grok-4",

Then 2) all models get to see each other's (anonymized) responses and they review and rank them, and then 3) a "Chairman LLM" gets all of that as context and produces the final response.

It's interesting to see the results from multiple models side by side on the same query, and even more amusingly, to read through their evaluation and ranking of each other's responses.

Quite often, the models are surprisingly willing to select another LLM's response as superior to their own, making this an interesting model evaluation strategy more generally. For example, reading book chapters together with my LLM Council today, the models consistently praise GPT 5.1 as the best and most insightful model, and consistently select Claude as the worst model, with the other models floating in between. But I'm not 100% convinced this aligns with my own qualitative assessment. For example, qualitatively I find GPT 5.1 a little too wordy and sprawled and Gemini 3 a bit more condensed and processed. Claude is too terse in this domain.

That said, there's probably a whole design space of the data flow of your LLM council. The construction of LLM ensembles seems under-explored.

I pushed the vibe coded app to
github.com/karpathy/llm-c…
if others would like to play. ty nano banana pro for fun header image for the repo
2
0
9
Replying to @isky
@isky your pup
2
0
1
Replying to @isky
@isky ????
1
0
0
Replying to @Sofi_Shvets
2
0
2
Replying to @daniel_dhawan
@daniel_dhawan btw this is a pic you get by googling Pumpkin Spice Latte Recipe
0
0
5
Built an app that edits pictures in a viral format in a single prompt.

drop a comment i will make you a pic with your avatar
👇
13
5
75
Replying to @vikhyatk
0
0
1
damn, this you?
It’s a good sign when competitors who started 3 months before you, copy you.

There’s a lot of “competition” in the vibe coding space. But surprisingly very little original thinking.

Now it’s time to just stay focused and build what we set out to build 6 months ago.
33
3
261
Introducing the first Vibe Coding IDE for iOS apps.
It makes creating apps feel like using CapCut.

🌃 Generate images, icons, backgrounds
🎶 Create music, sound effects
🤖 Create AI agents
👉 Available for preview in Rork
29
11
223
Replying to @vibecodeapp
@v_computer it is more like #15
1
0
5
Replying to @rauchg
is it possible to remove the requirement of "type": "module" for hono projects? When used in monorepos + pnpm we could not run it, so had to stay in CF.

or maybe allow to make trailingSlash: null, so i can use nextjs + hono, but getting full control of it.

Could not migrate one of oly
0
0
3
Let me explain what this AI photo editor app made in rork is doing
My friend got married. Instead of champagne, I gave her an app💍😅

As my wedding gift, I built Insieme - a photo-sharing app where friends can:
📸 Share photos
🙋 Add themselves into pics they missed
🧹 Remove people from the background (and make other custom edits)

I built the ith @rork + @cursor_ai, and used the @reve model for image edits.

Do you like the idea of gifting custom apps for life’s big moments?
5
2
31
some competitors have clearly nothing to do in life.

imagine seeing Rork is winning on design arena and spending some time to create spam votes to move Rork down.

pathetic.
2
2
19
Rork is on Product Hunt!
4
4
42
Rork started featuring projects.
Every once in a while we will select the best projects we find and feature them.

Let us know if you want to be featured.

Good luck!
1
1
29
Rork is #6 in Türkiye 🇹🇷🇹🇷🇹🇷
1
0
12
There are 136 countries in Apple Top Charts.

Rork is Top-5 in 47.
Rork is Top-10 in 91.
Rork is Top-15 in 109.
Rork is Top-20 in 120.

Rork is literally everywhere.
3
1
41
Extrapolate this
2
3
27
The history is written by builders,
the next builders are using @rork.
1
0
15
I love @tim_cook

I remember when reviewing an app took about a week if you were lucky.

Now it takes less than 24 hours, and if you have a hotfix, you can press 3 buttons and ship an update in an hour.
1
1
13
#2 in the US charts! @rork 🇺🇸🦅
20
4
114