11:56Alex Finn · Tutorial
Claude Sonnet 5 just dropped. I'm changing how I use AI
A 12-minute breakdown of Sonnet 5 benchmarks, a live ChatGPT 5.5 head-to-head, a two-model workflow for Claude Code, and leaked Fable 5 strings.
June 30thThe argument in one line.
ChatGPT 5.6 beats Fable 5 for coding not because it reasons better, but because its autonomous harness -- browser control, account signups, unattended task completion -- finishes real work reliably at a fraction of Fable 5's API cost.
Read if. Skip if.
READ IF YOU ARE…
- You run AI coding agents like Claude Code or Codex daily and want a second opinion on which model to route which task to.
- You've been shocked by API metering costs on a frontier reasoning model and want to know if there's a cheaper way to get similar output.
- You manage a coding project across a phone, tablet, and desktop and want a cleaner way to keep them in sync.
SKIP IF…
- You don't use AI coding assistants, so the model-vs-model comparison won't be actionable.
- You want a rigorous, benchmarked comparison -- this is one creator's subjective day-to-day impression, not a controlled test.
The full version, fast.
ChatGPT 5.6 launched as three tiers, Sol, Terra, and Luna, with seven thinking levels and a faster mode that folds OpenAI's Codex app directly into ChatGPT. Comparing it against Anthropic's Fable 5, the reviewer found Fable 5 still reasons better about large, unfamiliar codebases and high-level planning, but calls it unreasonably expensive after burning $500 of API credits in a day. ChatGPT 5.6 wins on execution: it finishes large tasks more eagerly and its harness -- controlling a browser, signing up for accounts, fetching API keys, running unattended tests -- inspires more confidence than Claude Code's. The recommended workflow: plan with Fable 5, execute with ChatGPT, and run every device off one always-on desktop hub.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Create a free account →Where the time goes.
00:00 – 00:34
01 · Cold open
Direct-to-camera hook promising a definitive verdict on ChatGPT 5.6 vs Fable 5 plus a workflow that changes how the viewer uses AI.
00:34 – 03:34
02 · What's new in ChatGPT 5.6
Three model sizes (Sol/Terra/Luna), seven thinking levels, a new fast mode, Codex folded into ChatGPT as 'remote sessions,' and praise for the iPad experience.
03:34 – 09:09
03 · Fable 5 vs ChatGPT 5.6 breakdown
A two-column comparison card walks through planning vs execution, price, UI, and harness quality; a usage-dashboard screenshot shows $507 of Fable 5 credits burned.
09:09 – 11:43
04 · The verdict: harness over raw intelligence
ChatGPT 5.6 is crowned the winner because its harness -- computer use, browser control, unattended task completion -- inspires more confidence than Claude Code's, despite Fable 5's planning edge.
11:43 – 14:35
05 · Setting up the multi-device hub
One always-on desktop becomes the main development machine; every other device remotely controls it. Each project gets one 'factory admin' chat plus separate per-feature chats.
14:35 – 16:55
06 · Split workflow: plan in Fable 5, execute in ChatGPT
Use Fable 5 inside Claude Code only for planning the spec, then hand it to ChatGPT/Codex for execution; demo shows ChatGPT autonomously testing a feature end-to-end in 24 minutes versus roughly an hour manually.
16:55 – 19:40
07 · The 'you do it' mindset and benchmark test
Every manual step (signups, API keys, file moves) now gets delegated to ChatGPT; declares it the new daily driver and runs a recurring first-person-shooter benchmark test.
Lines worth screenshotting.
- A creator burned $500 of API credits on Anthropic's Fable 5 model in a single 24-hour coding session, which he says makes it too expensive to use as a daily driver.
- ChatGPT 5.6 ships as three sizes -- Sol (smartest), Terra (mid), and Luna (efficient) -- with seven selectable thinking levels, up from far fewer previously.
- OpenAI folded its standalone Codex app into ChatGPT, rebranding mobile and iPad access as 'remote sessions.'
- Fable 5 still wins on planning, high-level reasoning, and understanding large unfamiliar codebases -- it grasped an entire existing 'software factory' setup almost instantly, while ChatGPT 5.6 needed manual nudging.
- ChatGPT 5.6 is judged the execution winner because it is more 'eager' -- it takes on large tasks and finishes them completely, a trait the creator says used to be Claude's signature strength.
- The tie-breaker for 'best model' wasn't raw intelligence -- it was harness reliability: computer use and browser control confident enough that the creator trusts it to finish a task unsupervised.
- The recommended workflow splits jobs by model: use the expensive reasoning model only for planning the spec, then hand the plan to the cheaper, more eager model for execution.
- Using an expensive frontier model as a coding subagent inside another tool wastes its best feature -- its own autonomous harness -- so the creator argues against nesting one model inside another's interface.
- A hub-and-spoke device setup -- one desktop as the always-on 'main development machine,' every phone, tablet, and laptop as a remote controller -- keeps a coding project's state from fragmenting across devices.
- Keeping a single high-level 'admin' chat per project for merges, PRs, and status, separate from per-feature chats, mirrors how a manager would delegate to a real engineering team.
- Letting the model handle every 'manual' step itself -- signing up for accounts, fetching API keys, connecting services -- cut an end-to-end feature test from about an hour to 24 minutes.
- Even after declaring a winner, the creator expects a next-generation model, rumored to already be in testing, within weeks, underscoring how fast the model-comparison content cycle turns over.
Route AI coding tasks by strength, not brand loyalty
MODEL ROUTING
The most capable model at planning isn't necessarily the best one to hand execution to, so treating harness reliability as a separate axis from raw intelligence changes which tool a task should go to.
01Cold open
- A confident, single-verdict framing at the top of a comparison video signals to the reader that the piece is opinion, not a neutral benchmark, and should be read that way.
02What's new in ChatGPT 5.6
- Splitting one model family into differently sized tiers (smartest, mid, efficient) lets a workflow use the cheap tier for routine work and reserve the expensive tier for hard reasoning.
- Folding a separate harness app into the main chat app removes a context-switch that used to cost users an extra download and login.
03Fable 5 vs ChatGPT 5.6 breakdown
- A model that reasons well about a large, unfamiliar codebase within seconds is a different skill than one that eagerly finishes every requested feature without pushback.
- API-metered usage on a frontier reasoning model can burn hundreds of dollars in a single day of coding, which prices it out of daily-driver use even when it's the smartest option available.
- Code quality being just as good between two models doesn't make them interchangeable -- eagerness to fully execute a task, not just output quality, decided the execution winner here.
04The verdict: harness over raw intelligence
- Autonomous harness behavior -- opening a browser, signing up for accounts, fetching API keys, running tests unattended -- matters as much as model intelligence for finishing real tasks.
- Choosing a daily-driver model is a different decision than choosing the smartest model; reliability and cost at scale outweigh peak capability for everyday use.
05Setting up the multi-device hub
- Designating one desktop as a persistent hub and treating every phone, tablet, and laptop as a remote controller into it keeps a coding project's state in one place instead of fragmented across devices.
- Keeping a single admin conversation per project for status, merges, and PRs, separate from feature-specific chats, mirrors how a manager would triage a real engineering team.
06Split workflow: plan in Fable 5, execute in ChatGPT
- Splitting a workflow into a planning model and a separate execution model lets you pay for expensive reasoning only where it earns its cost, and cheaper eager execution everywhere else.
- Nesting an expensive reasoning model as a subagent inside another tool's interface can waste its most valuable feature, its own autonomous harness, so the two are better used side by side, not stacked.
- Handing a model a full end-to-end test -- load a browser, run through the flow, repeat it several times, write a report -- turned a roughly hour-long manual task into 24 minutes of unattended work.
07The 'you do it' mindset and benchmark test
- Replacing 'I'll do that manual step myself' with 'you do it' for every account signup, key retrieval, or file move is what actually separates a supervised coding tool from an autonomous one.
- Running the same informal benchmark task across model generations is a cheap, repeatable way to sanity-check whether a new release is actually better, without needing formal evals.
Terms worth knowing.
- Harness
- The surrounding tool or interface (browser control, file access, autonomous task execution) that a coding model runs inside, distinct from the raw intelligence of the model itself.
- Sol / Terra / Luna
- The three size tiers of ChatGPT 5.6 in this comparison: Sol is the largest and smartest, Terra is mid-sized, and Luna is the fastest and most efficient.
- Remote sessions
- OpenAI's term for using ChatGPT or Codex from a phone or tablet to remotely control work happening on a separate always-on desktop machine.
- Factory admin chat
- A dedicated high-level conversation kept per project purely for status checks, merges, and pull-request management, separate from feature-building chats.
- Vibe coding
- Building software primarily by prompting AI models or agents rather than writing code by hand.
Things they pointed at.
15:02toolLinear
Lines you could clip.
04:18
“I spent $500 of credits in the last twenty four hours of Fable five.”
10:24
“The fact that you can give it a task and you know for certainty... it will be done for a reasonable price leads me to make ChatGPT 5.6 the winner.”
17:08
“Anything you would manually need to do before when vibe coding, I just tell ChatGPT to do it.”
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
metaphorstory
00:00It actually happens. ChatGPT 5.6 is out, and I am totally blown away.
00:06But is it better than Fable five? I have a definitive answer for you. In this video, I'll cover what ChatGPT 5.6 is, the improvements it made, do an in-depth breakdown against Fable five, show you who the clear winner is, and then show you an amazing workflow that will change how you use AI forever.
00:27This is the most important video you'll watch in a very long time. Make sure you lock in, and let's get into it. Days like today are the absolute best.
00:36If if you're into technology, congratulations. You are going to have a lot of fun right after this video. This is Chad GPT 5.6.
00:44I've had early access. I've been using it a ton. So all the people at GONA comments go, how'd you get a video out so quick?
00:50What the hell? What's going on? Shut up.
00:51I've been using it. Now let's get into it. Chad GPT 5.6, three models, Luna, Soul, Terra.
00:56The one we're gonna talk about today is Soul. That's their biggest model. That's the one everyone goes, oh, how's it compared to Fable?
01:02Is it better than Fable? That's the one we're gonna be talking about is Soul. So Soul's the smartest.
01:07Terra's kind in the medium. Luna is the efficient one. You chop GPT mini.
01:11So Soul's the one we care. I care about brains. So let's talk about that.
01:15They also have many different thinking levels now. They have, like, seven different thinking levels. Ultra is the one I'm using.
01:22The reason why we can use the smartest CHAT GPT model with the highest level thinking is because ChadGBT is so kind when it comes to usage.
01:32It gives you a ton of usage, so you don't have to be scared like you are with Anthropic of using the highest thinking levels. Then two prompts later, you have no more usage. ChadGBT, you can use the highest thinking levels and still get a ton of usage out of it.
01:46They have a fast mode, which is, in my opinion, revolutionary. I'll tell you, you're getting fable level thinking on a really, really fast speed.
01:57One challenge I've had with Fable over the last couple weeks is it's just slow. It's a slow model. But Chad GPT five six Soul Ultra on fast mode, lightning fast.
02:09It is the best model in some ways. We'll get into the Fable comparison in just a second. Stick around me here.
02:14We go in-depth. Who's the winner? I hand a trophy out.
02:17It is the best model in some ways, but it is without a doubt the best harness codex, which we'll also talk about that in a second. It's gone.
02:27It's the best harness at the best price at the best speed.
02:32Some also also some changes announced today. We'll go into in-depth breakdowns on subsequent videos.
02:39Let me know what you want me to do most. ChatGBT works, some of these other things. Codex app, no more.
02:44It's gone. They combined it with the ChatGBT app, which I actually believe is a very smart move.
02:50Codex was the best harness. They're bringing all those nice features over to the ChatGBT app.
02:57It's all in one. I didn't like the separate apps. I'm a fan of this move.
03:00Codex is no more. It is changed on mobile as well. So if you use Codex on mobile, if you use Codex on iPad, which I highly, highly recommend you do, it's now called remote sessions.
03:12They really want you using this in a very specific way on mobile, which I'll talk about a little bit in a second as well. I actually think the iPad experience now is, like, the best experience. I I use the iPad a ton.
03:23I'll talk about that too. But the iPad experience for chat I don't even think they're, like, trying to make it good.
03:28It actually is very good. I'll talk about why I think it's awesome too. But let's do the comparison everyone and their mothers is waiting for.
03:36ChatGPT five six with Fable five, which is the winner? Well, I'm gonna show you the comparison between the two, what the differences are, and then I'll tell you the winner once we go through this.
03:46We'll make this quick. I have been using the absolute booty load out of Fable five last week. I don't even know what booty load means.
03:52I just dropped that. As you can see here, I hit extra usage yesterday. I spent $500 of credits in the last twenty four hours of Fable five.
04:01I've been using the absolute hell out of it. So I am in-depth in the weeds on Fable five. I know how it works inside and out.
04:09How does it compare to 5.6? Well, let's take a look at this here. Fable five is still better for planning in high level thinking and understanding of large code bases.
04:20It is the best when it comes specifically to that. I'll give you an example.
04:24I have, like, an entire software factory I've built out inside of Codex and ClaudeCode that allows me to kinda autonomously build a lot of software. It's very complex, in-depth. If want a video on that, let me know down below.
04:37I'll do a whole video on loops and software factory if you're interested. But here's the thing. When I gave that to Fable five, I said, hey.
04:44Look at my entire software factory and understand how it works. It understood it pretty much instantly. It was able to pick it up and start doing work in the factory.
04:53Like, it was been like, it's been there for years. When I handed the software factory over to chat GBT five point six initially when I got early access, I'm gonna be honest, it did not pick it up immediately. It took a while.
05:05I had to kinda nudge it in different directions until it finally grasped how it worked. So when it comes to understanding large code bases, being a high level wise thinker, understanding from, a CEO level what's going on, I would say Fable five's better.
05:23But now let's talk about actual execution, actual building. 5.6 is the winner.
05:295.6 is the winner for many, many different reasons. The code quality is just as good as Fable five. The code is putting on is just as good, but it is more eager.
05:39So here where ChadGPT 5.6 is more eager than Fable. So it will get tasks done.
05:46You can give it major, major tasks, and it will grab it. It will get it all done.
05:51It will add a whole bunch of extra features to it. It'll be fantastic at that. This was Claude's strength for a long time.
05:57That kind of eagerness of getting a ton done was Claude's strength for a very long time. ChadGBT has overtaken that. So when it comes to execution, when it comes to actual building, five six is the winner.
06:09Now after this comparison we're going through, I'm gonna actually show you the workflow. Everyone that mothers is going on Twitter saying, oh, use ChadGBT as a sub agent inside of Claude Coat.
06:19No. That's not the best way to do it. I'll show you the best way to do it a second.
06:22You'll learn the best workflow once we go through this, once I declare the winner between ChadGBT 5.6 and Fable five. Workflow coming right after this. Let's talk about price, and this is gonna be a deal breaker for most people.
06:33Fable five, unreasonably expensive. So expensive, I actually think this model will die.
06:39I think Fable five is the smartest AI model, potentially the greatest technology ever made. I think it's dead in the water. I was shocked when I started going into API pricing, and I, within a few hours, used hundreds of dollars of credits.
06:54That's absolutely insane. No normal person is going to wanna pay that price. And then you even look at the enterprise right now, companies are cutting out frontier models and using things like GLM 5.2 because costs are getting so high.
07:09I don't know who Fable five is for. It's absolutely genius level. I wasn't prepared to tell everyone, hey.
07:15You need to be paying API pricing for it. And the moment I started paying API pricing for it, I'm like, wait a second. No.
07:21No. No. No.
07:21No. Done with this. Done with Fable.
07:24Unreasonably expensive. It is so unreasonable, in fact, that I believe they're going to push up Opus five.
07:31I think in the next week, Opus five comes out. I believe I don't think it's Fable level. I believe it's very close to Fable level, but at way more reasonable prices.
07:41I just don't see how Fable fits into the picture with what I'm about to show you with Chad GPT. No one in their right mind, unless you're just outrageously rich, is gonna be spending $500 a day on credits.
07:51And this was even like a full day of coding. This was like an afternoon, like 1PM to 8PM of coding.
07:56$500. That is crazy. ChadGPT 5.6, it's more affordable.
08:02Right? It's way more affordable than Fable five. It is still costly.
08:06I have never I'm gonna be honest with you, and I do this all day. I've never hit my limits with ChadGPT before. That's never happened for me.
08:14Five five, everything before that I've never hit my limits. I hit my limits for the first time yesterday using ChadGBT 5.6.
08:23That's never happened before. So it is more expensive than the other ChadGBT models.
08:29I did hit my five hour window. I didn't come close to the weekly window. Five hour window, though, I hit the limit.
08:34I I started having to use API pricing. I was using Soul at ultra level. So, like, if you go super hardcore on their most expensive model, you could hit limits.
08:46But at the end of the day, it is way more affordable and reasonable than Fable five. UI was always Chad GBT's weakness. It's now good.
08:54It's good enough at UI now. It it's certainly improved it. It's good enough.
08:58It's not a concern now where you need to jump to Claude every single time. Claude is better at UI still. Here's where I actually think the winner comes in, and this is why I am going to declare ChadGBT 5.6 the winner.
09:12The harness around ChadGBT is so much better. It's so much better.
09:17The computer use is significantly better. The browser use is significantly better. Claude has made jumps here lately, But ChadGBT, the harness doing things for you, signing up for websites, getting API keys, I've never been more confident when giving a model a task that it will do literally whatever it takes to get it done.
09:39It will pop open the browser. It will control your computer. It will do whatever it needs in order to get that task done.
09:47And that is a feeling and emotion that I've never had with Claude inside their harness, which for me leads me to crown ChadGBT 5.6 the winner because that feeling of knowing when you give a task to a model, knowing it will get done is an invaluable feeling.
10:08Even though there's some weaknesses when it comes to planning, high level thinking, figuring very large code bases out that it doesn't really match up with Fableon, the fact that you can give it a task and you know for certainty at the end, it will be done for a reasonable price leads me to make ChadGBT 5.6 the winner.
10:28The question becomes also little things here. I'll do a video. Let me know down below.
10:32Do you want a full video on Hermes using ChadGBT 5.6 as well? I'll do that. But this leads me to the next section of this video, which is the workflow.
10:42How do you actually use it? We talked about what is better, how it compares to Fable five, what the new changes are. How do you actually use it.
10:51Everyone and their mother's been telling you if you pay attention to Twitter, if you're in the Twitter bubble, oh, set up chat gbt five point six as a sub agent inside Claude Cove. No.
11:00Do not do that. I will show you the best way to do this. And before we do that real quick, if you learned anything so far, leave a like down below, subscribe, turn on notifications.
11:10All I do is make the absolute stone cold best content about AI on the entire Internet. And if you wanna work with me live, I do live boot camps every single Friday in the Vibe Coding Academy.
11:23It's the number one community for people building with AI on planet Earth. We're also doing a thousand dollar hackathon right now. It ends in two weeks.
11:31You can still join it, get in, build something cool, submit it, you go in a thousand dollars. We're doing it every single month. Join the Vibe Coding Academy.
11:39Link down below for that. Let's get into the workflow now. So this is the Codex app.
11:44I'm using Codex app for now because I am filming this right before the model comes out. So I'm still using the Codex app before it's combined with Chad GBT. But here's how you wanna use this.
11:54Here's how you wanna use Chad GBT 5.6. First, we'll talk about from the harness level. You want to choose one computer or device you own to be kind of the head development machine.
12:06The reason why you wanna do this is the remote control capabilities of Chad GPT are incredible. You can control Chad GPT from any device you have, iPad, iPhone, any of your computers, Mac mini.
12:19So what you wanna do is choose your kind of main developer machine. This should be a desktop device.
12:26I'm using one of my Mac studios, my Mac Studio One. I have three of them, not to flex on you, but kind of a flex.
12:33I'm using my first one as my main developer machine. Now every other device you have . Here's how you wanna think about it.
12:42So you have your main development machine where you're using ChatGPT or codecs, and then every other device you own will control that. So if you have a desktop and you have a laptop, your laptop controls it, your iPad, your iPhone, everything controls this one place. And the reason why you wanna do it this way is so that you have all your code in one place.
13:02Instead of having some features on your laptop, your desktop, some features on your iPhone, your iPad, it all goes to one place. And because the remote control functionality is so good on the ChatGPT and Codex apps, you can easily control from all your different devices. So choose a device.
13:18Make sure it's a desktop if you have one. The reason being is you can leave your desktop on 247365. Laptops, you gotta close it, puts it in sleep mode, different things like that.
13:29So make sure you have kind of a desktop device, a Mac studio, a gaming computer. Whatever you got, make that your main computer. Even if you just have, like, an old dusty laptop in your closet that you can put on twenty four seven mode and plug it in, use that.
13:42Something that can be up twenty four seven, three sixty five. So from there, you have your projects and you have your chats inside of each of the projects. I'm going over kinda high level tips in here.
13:52We can do a deep dive building out a full app end to end on a subsequent video if you want. Let me know down below. But what we're gonna do is you wanna kinda have your factory admin chat in each project.
14:03This is the high level chat where you can be like, okay. Merge the code, create a PR, do this, where it's kinda managing your entire factory. And then every other chat from there is gonna be building out specific features for that project.
14:18So you have your main admin chat where you can go in from any of your devices and say, hey. What's the status? What's uncommitted code we have?
14:25What PRs do we have to merge? Things like that. It's kinda your high level admin.
14:29And then every other chat is for specific features you're building out. Now we said earlier, Fable five is a better planning model, and I I stand by that.
14:40So what you wanna do is you want to use Fable five for planning. So if you have a project, you open it up inside Claude code, you go into Fable five mode. I'd go on high.
14:51If you go on ultra code or higher, you spend hundreds of dollars from a single prompt. Just use high. When you're doing any sort of planning, you'll come into Claude code.
15:01You will ask for the plan. Hey. This is what I'm thinking about building.
15:04Please build out a very large spec. If you've been paying attention to all my videos, you know I talk about linear and how I use linear in my entire process. So if you're using linear, have it build the spec out in linear.
15:14If not, totally fine to say, hey, build a markdown file. Then what you will do is you will take that markdown file, go into your projects, drop it in, and say, here's the plan, build this out for me.
15:27Now a lot of people are telling you to use a ChadGBT sub agent inside of Claude Co. To do execution.
15:33I don't like that at all because when you do that, you're missing out on all the amazing features of the ChadGBT harness.
15:41The Chad GBT harness does so many things so well for you. So for instance, I'm building out my app, Henry Intelligent Machines. Link down below if you wanna get an early preview of that.
15:52And I'm having it test some of the action generation inside the app. In Henry, you launch a venture, and then it generates a bunch of actions that you need to do to push your business forward. I wanted to test that.
16:03Typically, when testing that, you would need to open up a browser, go through the flow yourself, click on things yourself. But all I do is I go in and I say, hey, test this out for me. Load it up into a browser.
16:16Walk through it for me. Do it five times and then build a report on that. And that is what ChadGPT does.
16:23It opens up its own browser and runs all the servers. It gets whatever API keys it needs. It plugs them in, and it runs through everything.
16:32So it said what I tested actually opened it up, started a venture, did a bunch of different actions, tested everything out. It did all the testing for me. Something that would have typically taken me an hour before took twenty four minutes for chat gbt while I did other things.
16:48So your workflow is get the planning from Claude code, move it over to chat GPT, and then you give the actions to chat GPT. Here is the mindset I have.
17:01Anything you would manually need to do before when vibe coding, I just tell ChadGBT to do it.
17:09Anything it is. It doesn't matter what it is. If they say, hey.
17:11Sign up for this account. I go, no. You do it.
17:13If they say, hey. Can you grab this API key? I say, no.
17:16You do it. If they say, hey. Can you connect these two things together, move this file over here?
17:20I say, no. You do it. ChadGBT literally 100% of the time, I have said, hey.
17:26No. You do it. It's figured out how to do it.
17:29That's why this harness is so good. When it comes to computer use, when it comes to browser use, it is absolutely next level. It is bar none, not comparable to Claude code whatsoever.
17:39That's why the workflow is get your plans from Claude code, move them over to Chad GPT codecs, and then you say, no. You do this. You handle this.
17:48And it will get done, and it will save you tons of time. Now while you're in here, I do Soul Ultra for everything. It's the stone cold best.
17:56If you're on the $200 a month plan, you can get away with using fast for a lot of things. Again, I hit my limits using fast here for the first time my entire life in chat gbt. So maybe you don't use fast quite as much, but if you're on Soul Ultra on normal speeds, you can use up your $200 plan.
18:12Like, you you're not gonna hit your limits. You're gonna have plenty of usage, which I cannot say for Fable because I use the plan up in, a day.
18:20And then when I went into API pricing, it was outrageous out of control. 5.6 is the winner.
18:24It is my new daily driver. It is what you need to be using for pretty much everything outside of high level understanding, planning, and some kind of UI tweaks if need be.
18:34It's still pretty good at UI. By the way, for the hardcore Alex Finn fans who know all about the world famous Alex Finn benchmark, yes, it blew away the benchmarks. This is the first person shooter test.
18:45It is absolutely incredible. It has it has crits, dynamic enemies. It has great explosions.
18:51It has great waves. The graphics are fantastic. The sounds are amazing.
18:55I don't know if you can hear the sounds or not. The power ups are incredible. Like, there's, like, 10 different dynamic enemies.
19:00It's incredible. Chad GPT 5.6 won the benchmark test. I am really, really impressed.
19:06And what's incredible is over the next few weeks, we're actually expecting GPT six. We're expecting the next level. The the rumors I am hearing are that it's coming very soon.
19:16They're already done training it. They're just doing testing now, so expect more videos like this very, very soon.
19:23If you learn anything at all, likes, subscribes, turn notifications. Again, live boot camp tomorrow, thousand dollar hackathon next couple weeks.
19:32Join the Vibe Coding Academy. Link down below. I love you all.
19:35It's an honor to teach you these things. It's an honor to work with you. We will talk soon.
The Hook
The bait, then the rug-pull.
ChatGPT 5.6 is out, and within a day its creator had already burned $500 in Fable 5 API credits trying to keep up -- so which model actually wins, and does it matter more than the workflow wrapped around it?
Named ideas worth stealing.
00:36list
Sol / Terra / Luna model tiers
- Sol - smartest, flagship reasoning tier
- Terra - mid-sized tier
- Luna - fastest, most efficient tier
ChatGPT 5.6 ships as three sizes so routine work can run on a cheap tier while hard reasoning goes to the flagship.
Steal forany AI workflow that needs to route tasks by difficulty and cost
12:39model
Hub-and-spoke device setup
- iPhone
- iPad
- Laptop
- Other computers
- MAIN DESKTOP
One always-on desktop is the persistent execution hub; every other device is a node that remotely controls it.
Steal forkeeping a coding project's state in one place across multiple personal devices
13:53concept
Factory admin chat pattern
One high-level chat per project handles status, merges, and PRs; every other chat is scoped to a single feature.
Steal formanaging multiple parallel AI coding threads without losing track of project state
How they asked for the click.
VERBAL ASK
11:07product
“leave a like down below, subscribe, turn on notifications... join the Vibe Coding Academy... thousand dollar hackathon”
Soft mid-video pitch for his paid Skool community inserted right before the workflow section begins, repeated again in the final seconds.
MENTIONED ON CAMERA
FROM THE DESCRIPTION
PRIMARY CTAWhere the creator wants you to go next.
OTHER LINKSAlso linked in the description.
Visual structure at a glance.
