@anshuc profile picture
Anshu @anshuc
Joined June 2009
  • Posts 2,026
  • Following 1,060
  • Followers 3,507
304 Photos and videos
Pinned Post
Anshu
 @anshuc
Jun 10
Some of you said it still looks like AI slop, and I agreed, so I told Fable to "100x the design." Here's what it did, and all the details on how.
0:20
Anshu
 @anshuc
Jun 10
x.com/i/article/206480015066…
74
136
2,424
719,067
Update: If you want to add these kinds of visuals to your app, here's my prompt and all the details x.com/anshuc/status/20739658…
Anshu
 @anshuc
10h
Here's my workflow & prompt for building these mind-blowing cinematic transitions with Claude Fable! I spent many days and $100 in API credits figuring all this out. You can do it in half an hour for $2 with this guide. Enjoy!
0:09
1
299
Here it is! x.com/anshuc/status/20739658…
Anshu
 @anshuc
10h
Here's my workflow & prompt for building these mind-blowing cinematic transitions with Claude Fable! I spent many days and $100 in API credits figuring all this out. You can do it in half an hour for $2 with this guide. Enjoy!
0:09
99
Here's my workflow & prompt for building these mind-blowing cinematic transitions with Claude Fable! I spent many days and $100 in API credits figuring all this out. You can do it in half an hour for $2 with this guide. Enjoy!
0:09
Anshu
 @anshuc
10h
x.com/i/article/207323921296…
3
6
42
2,960
x.com/i/article/207323921296…
8
4
64
3,888
For everyone asking, my how-to article for these insane cinematic visuals is almost done, and I’m super excited about it! I’ve boiled it down to one prompt, and cut the costs a ton. You’re gonna love it :) Hopefully tomorrow, keep an eye out!
0:49
9
3
99
6,620
Hope this helps someone. I’m incredibly grateful for all of you who read and share my posts, ask questions, and give feedback. I’ll do my best to pay it back by sharing whatever knowledge I have!
7
365
3/ Forget replies. Aim unreasonably high Common advice is to be a reply guy when you’re a small account, since you don’t have your own audience. I say it’s a waste of time. I reply plenty, but it’s to engage with people I like, build relationships, share ideas. Not to borrow their audience. Everyone can tell when you’re doing the latter. Yes you can grind replies and add a few followers every day. You can also go massively viral once and add 1000 followers instantly, while learning the exact formula for success, without needing a bigger account to carry you. Another underrated point: your own viral posts are an extremely valuable library you can reuse and build on. Much harder to do that with replies. You need to aim higher than you think. If you have 100 followers and one post gets 5k views, don’t settle for the small win. Get out of the “it’s a marathon, not a race” mindset. Screw that. I hate marathons. Sprint to the finish line. A 100 follower account can get 1M views with incredibly engaging content. If you got 5k views, you’re on the right track but the content’s not good enough. Iterate and aim higher. This will get you growth in days that would otherwise take months.
1
1
414
2/ Write content relative to your account size Nobody talks about this. I didn’t understand it for the longest time. I knew my content was good because I saw bigger accounts posting similar content with amazing results. So why wasn’t it taking off? The performance of your posts depends on how engaging the content is, times your audience. Elon can post “Concerning” and get 10M views and 20k likes. You can’t. His reputation and distribution are carrying the post. You don’t have those, so your post needs to carry itself. Your job is to post content that people want to share. To grow your network, your content needs to reach outside your network. I do this by posting things that are beautiful and provoke a response. People want to comment/repost/bookmark because it looks cool and teaches them something. If you have similar talents, just copy my playbook! If not, find the recipes that work for you (point 1). Copying others is a waste of time; your distribution and reputation are different from theirs.
2
2
280
1/ Don't "just post", post in search of winning formulas I spent a long time posting into the void. The "just post, anon" strategy works for some, but they’re people who naturally write what a lot of people want to read. Great, but what about the rest of us? Praying the algorithm will randomly make you viral is a trap. You need to be methodical. Each post should have a hypothesis on why it'll go viral. If it doesn't, you learn something that gets you closer to the goal. There are no failures. Here’s my story: I shared demos, AI takes, memes, etc., for months with no results. When Claude Fable dropped, I posted an app demo that looked especially cool. I went to bed, woke up, and had 500k views and a ton of likes/replies. I learned a few key things: being on trend with Fable helped. Strong visuals helped. There’s clearly an audience for building beautiful apps. It’s easy to see that as dumb luck, but it’s not. Looking at my old demos, I now see exactly why they failed: not visually engaging enough, not topical enough, not educational enough. Meanwhile, later ones built on the same recipe had even better performance. The recipe is what matters. To find your recipe, you need to iterate rapidly without repeating mistakes. If you make each post an experiment, virality becomes a science. Once you find a winner, start changing one variable at a time and testing alternatives. This lets you optimize your strategy and grow into adjacent audiences. That’s the long game, but it's consistent and predictable, not random.
2
4
427
In 3 weeks, I added 3000 followers, 7M impressions, and got my first payout! After 1 year of getting nowhere at 200 followers, I’ve finally learned what works. Most advice is too slow or plain wrong. If you feel stuck, READ THIS. You can 10x your account by August 🧵
9
29
2,564
Lots of replies in the quote post said Kickbacks was a scam. Well, the money just hit my bank account. My Claude subscription made me $100! Never doubted you @andrewmccalip :)
Anshu
 @anshuc
Jun 11
this thing is actually going to make my claude subscription profitable at this rate
5
7,063
One takeaway from playing with Fable: the clearer the image in your head, the better the results People ask me for the prompts, and I share them, but they’re the output, not the input. My best work comes from going on a walk or closing my eyes and trying to picture every detail of the app, every transition, as vividly as I can. I often dictate or write notes while doing so. I try to be intentional about what I want to see vs where I want the model to fill in gaps and surprise me. That’s what turns into the prompts A lot of it is iterative too. I see the first result and it inspires something - what if the fish felt alive? what if the dishes were 3D? That’s why my preferred approach is minimal tools, since they add friction or constraints. I don’t usually make sketches. I just try to hold on to what’s in my head and let my imagination guide me
Anshu
 @anshuc
Jul 3
Fable is pure magic. I wanted a beautiful app to explore ocean wildlife. Fable built this in an hour. It generated videos with Seedance and carefully synchronized them to make these absolutely insane transitions. I've never seen anything like this in an app. Unreal.
0:49
2
1
26
2,891
Fable with Cursor on iOS is pretty sweet! Since it has its own VM, it can actually test projects end to end and send me demo videos as proof I’m Fabling while my flight is taking off! (And yes, this space game is not dead, hang tight)
0:15
1
14
1,853
I’ll share my workflow and prompts for this soon, and stick it on the App Store for free - stay tuned! If anyone’s got requests for apps you want to see me overdesign the hell out of with Fable, let me know. I’ve got more ideas cooking too :)
Anshu
 @anshuc
Jul 3
Fable is pure magic. I wanted a beautiful app to explore ocean wildlife. Fable built this in an hour. It generated videos with Seedance and carefully synchronized them to make these absolutely insane transitions. I've never seen anything like this in an app. Unreal.
0:49
5
53
5,715
This took a few iterations and nudges, but the entire app concept was Fable's. Fable designed all the visuals and assets using Gemini and Seedance. Oh, and if you don't get what makes Fable special, here's what Sonnet 5 did with the same prompt…
3
2
80
9,538
Fable is pure magic. I wanted a beautiful app to explore ocean wildlife. Fable built this in an hour. It generated videos with Seedance and carefully synchronized them to make these absolutely insane transitions. I've never seen anything like this in an app. Unreal.
0:49
108
158
2,998
263,934
Getting a lot of questions about the 3D asset workflow; here’s my post on that:
Anshu
 @anshuc
Jun 23
Everyone's sleeping on image-to-3D AI models. They can make your app look incredibly unique, with just a little effort. Here's how. This is my calorie tracker, built in a week with nothing but prompting. Just Claude Code a couple APIs. The visuals are all AI-generated. I'll be sharing the full workflow all the crazy technical stuff Claude and I did to make this work, so nobody has to struggle through it like me. Deep dive coming soon! Till then, this is the high-level idea: 1. Get a clean image of the food (or whatever your asset is) - In my app, the user describes foods via text, or attaches images (or both) - If text, an LLM extracts the food description and formats it into a specific prompt I tuned for this design, and we generate an image using Z-Image Turbo through @fal - If image, we do the same thing but with FLUX.2 [dev] to edit the user image into our reference design - Originally, both used Google Nano Banana, but switching to open models cut costs and latency a ton 2. Gaussian splatting (2D image → 3D model) - I tried various 2D-to-3D options on @fal and ended up with TripoSplat as my preferred balance of speed, cost, latency; this turns an image into a 3D model that looks super high quality (link below) - The app displays the 2D image while our backend generates the 3D splat - We "groom" the splat to reduce size and load time by culling low-opacity/scale points 3. Render efficiently on device Originally, it looked great but ran at 10 FPS. Getting to 120 FPS was a crazy journey. TL;DR: - SwiftUI had to go; it forced us to render each asset in independent MTKViews, which wasn't workable - Instead, we composite every dish into one full-bleed CAMetalLayer using MetalSplatter (link below) - We had to make some optimizations within MetalSplatter's code too, to reduce the overhead of sorting points per render Then I added some finishing touches like the subtle rotation and parallax as they move around. I think it turned out pretty cool :) Overall, this took some effort, but we still got it done in less than a day. Hopefully your agent can follow in the footsteps of mine and do it much faster. Keep an eye out for the bigger writeup, which'll give your agent everything it needs. If you have any questions, drop em below!
0:15
8
3,733
finally, how I missed this model
Anshu
 @anshuc
Jun 30
Good news: Sonnet 5 will NOT be export controlled
4
48
7,162
Fable is FINALLY back! Here’s exactly how I used Fable to build this app that got 1M views on X. You have everything you need now - no more excuses, go build something incredible!
0:32
Anshu
 @anshuc
Jun 10
x.com/i/article/206480015066…
48
136
1,800
567,749
Pretty impressed by new Siri being able to do this (and also understand my god awful typing)
1
22
2,085
Load more