Big Idea

The argument in one line.

Claude Fable 5 changes the job from step-by-step direction to goal-setting: give the model an ambitious end state, let it run a loop, and check the outcome rather than each individual action.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…

You are already using Claude Code with Opus or an earlier model and want to understand what changes with Fable.
You want a live demo of the /goal loop pattern for autonomous multi-feature app builds.
You are deciding whether the price jump from Opus to Fable is worth it before the June 22 subscription cutoff.

SKIP IF…

You have never opened Claude Code — this assumes CLI familiarity.
You want a deep technical or safety explainer; this is a practical workflow walkthrough, not an architecture analysis.

TL;DR

The full version, fast.

Claude Fable 5 (also called Mythos) tops every AI benchmark and is now accessible via Claude Code CLI. The core argument: the mental model shift with Fable is moving from checking whether each step is correct to setting an ambitious goal and validating the final result. The practical workflow demonstrated is advanced plan mode (a prompt that asks Claude to interrogate your requirements before building anything) followed by a /goal command that runs Claude in an autonomous loop until stated success criteria are met. The model is included in subscriptions only until June 22, 2026, then becomes API-only at double Opus pricing.

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 →

Chapters

Where the time goes.

00:00 – 00:57

01 · Hook + benchmarks

Claude Fable 5 / Mythos released; dominates every benchmark; available in Claude Code CLI, desktop, chat, and co-work.

00:57 – 02:09

02 · Mindset shift

From checking whether each step is correct to checking whether the right goal was set; loops introduced as the mechanism.

02:09 – 03:23

03 · Anthropic's three tips

Treat Claude as a thought partner, give goals not steps, be more ambitious — the official guidance at launch.

03:23 – 04:56

04 · Setup in Claude Code CLI

Ghosty terminal, /model claude-fable-5, shift+tab for auto mode, effort set to high.

04:56 – 06:36

05 · Advanced plan mode

Prompt pattern that asks Claude to interrogate requirements before building; applied to designing a full productivity app.

06:36 – 08:00

06 · /goal loop kickoff

Answering Claude's questions, generating the goal command, and launching the autonomous build run.

08:00 – 10:46

07 · App demo

Reviewing the finished app: task list, Kanban, calendar, notes, Pomodoro timer, and snake game — all functional, zero claimed bugs.

10:46 – 11:58

08 · Advanced loops

/loop command paired with Linear integration; autonomous agents that poll external tools on a schedule.

11:58 – 12:57

09 · Pricing + CTA

Subscription window closes June 22; then API-only at 2x Opus price. Subscribe/notifications ask.

Atomic Insights

Lines worth screenshotting.

The mental model shift with Fable is from checking each step to checking the outcome — the model handles correctness, you handle direction.
Running /model claude-fable-5 in Claude Code grants access to Fable even if your model picker has not updated yet.
Shift+tab in Claude Code activates auto mode, eliminating per-action permission prompts — the recommended setting for autonomous Fable runs.
Set effort to high rather than extra-high for Fable; it torches tokens aggressively and extra-high is overkill for most tasks.
Advanced plan mode front-loads all requirement discovery before a single line of code is written — Claude interrogates you instead of assuming.
The /goal command wraps a build in a loop where Claude self-evaluates against stated success criteria and keeps working until they are met.
Fable is available in Claude subscriptions only until June 22, 2026 — after that it is API-only at double Opus pricing.
One /goal prompt built a to-do list, Kanban board, calendar, notes app, Pomodoro timer, and a snake game simultaneously with zero stated bugs.
The /loop command can poll an external tool like Linear on a set interval, creating an agent that independently picks up and completes coding tasks.
Benchmark dominance is real, but the live app build is the more credible validation — working software beats leaderboard numbers.

Takeaway

What Fable changes about working with Claude Code

MINDSET SHIFT

Fable does not make Claude faster at following instructions — it makes granular instructions unnecessary.

02Mindset shift

Shift from checking each step to checking the outcome: set success criteria up front, then evaluate the final result rather than supervising the process.

03Anthropic's three tips

Use a pre-planning prompt before any build to ask Claude to interrogate your requirements — it surfaces edge cases you would otherwise discover mid-build.

04Setup in Claude Code CLI

Set effort to high rather than extra-high for Fable; the model thinks hard by default and extra-high burns tokens without proportional gains for most tasks.
Auto mode removes per-action permission prompts and is appropriate for Fable because its judgment is reliable enough that constant confirmation overhead costs more than it protects.

07App demo

The /goal loop keeps Claude working until its own evaluation confirms the goal is met — no babysitting, no re-prompting.

08Advanced loops

Pairing /loop with an external project management tool (Linear, Jira) creates an agent that independently picks up and completes coding tasks on a schedule.

09Pricing + CTA

The June 22 subscription cutoff is a real deadline: stress-test Fable now while it is included, before it becomes API-only spend at double Opus pricing.

Glossary

Terms worth knowing.

Claude Fable 5 / Claude Mythos: Anthropic's frontier model released in June 2026. Mythos was the internal codename; Fable 5 is the public name. Previously withheld from general access due to safety concerns before this release.
Auto mode: A Claude Code setting (shift+tab to toggle) that removes per-action permission prompts, allowing Claude to execute tool calls without pausing for approval.
Effort level: A Claude Code parameter (/effort) controlling how deeply the model reasons before responding. Ranges from low to extra-high; higher settings consume more tokens.
/goal command: A Claude Code command that frames a task as an autonomous loop. Claude self-evaluates its output against success criteria embedded in the goal prompt and continues working until they are met.
/loop command: A Claude Code command that schedules a recurring autonomous check-in at a defined interval, useful for polling external tools like Linear for new tasks.
Advanced plan mode: A user-defined prompt pattern (not a native command) that instructs Claude to ask all clarifying questions before generating any plan or writing any code.

Resources

Things they pointed at.

03:23toolGhosty

11:00toolLinear

00:00toolClaude Code CLI

Quotables

Lines you could clip.

01:18

“Is Claude doing the work right? vs Is Claude doing the right work?”

Clean one-liner that captures the entire mental model shift — no setup needed.→ TikTok hook↗ Tweet quote

04:37

“You give it a massive goal, and it attacks that massive goal.”

Tight, quotable, standalone.→ IG reel cold open↗ Tweet quote

12:17

“This might be the beginning of the end of subsidized subscriptions.”

Contrarian prediction with real stakes — AI economics clip.→ newsletter pull-quote↗ Tweet quote

The Script

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.

00:00I cannot believe it. The biggest model release of the entire year, maybe of all time, happened today just now. We are about to find out if it lives up to the hype.

00:11Claude Fable five, aka Claude Mythos. That's right.

00:16Anthropic has released Claude Mythos five. It is the safe version of Claude Mythos. And as you can see here, it not only wins every single AI benchmark ever made, it dominates every AI benchmark ever made.

00:32But are benchmarks real? Is it as good as all the benchmarks you see here? Because a lot of time, benchmarks are a little fake.

00:38We're about to find out. For those who don't remember, Claude Mythos was the model that Anthropic announced a few months ago, but they said, hey. You normies, you can't use it a little too dangerous for you.

00:50You get your hands on this. The entire world is ending. It is available in all Claude platforms, so you can use in Claude code CLI in the desktop version as well as in the chat and in co work.

01:01I'm gonna show you how to set it up in a second, show you how to get more out of it. We're about to go through the keys of Claude Fable, so how you treat it different than Opus. So before we get into actually using the model, here's how you need to think differently about Claude Fable five.

01:18Before you ask yourself, is Claude doing the work right? You tell Claude Co. To do something.

01:23You tell Claude to do something, and then you go and you check to see if it did the right thing. You test it. Now it's different.

01:29This is a subtle shift in mentality. Now you are saying, is Claude doing the right work? You can now be confident Claude is building things the right way.

01:40It's not building buggy code. Now you just need to make sure it's doing the right work. How do you do that?

01:46Well, through loops. And I'm gonna go through loops in a second. This has been the talk of the town for the past week.

01:52I'll show you how to use loops using a few different methods inside Claude code in just a second as well. This basically gives Claude the ability to work autonomously nonstop, right, for days and days and days on end.

02:05We'll go through that very shortly. So here are the three tips Anthropic gave when it came to using Fable five.

02:12Treat Claude like a thought partner. This is quite literally superintelligence.

02:18And what people were doing before with Claude was they'd begin work, and they'd say, okay, Claude. Do this next. Do this next.

02:24Do this next. Not anymore. Now you need to treat it as an equal, as an equal collaborator on what you're building.

02:31So before you build something, you need to go, hey. What are your thoughts on this? How would you change this?

02:36How would you do it different? How would you approach this? Would you do something completely different?

02:41You treat it as an equal partner in what you're building. Two, you give Claude goals. Instead of telling Claude what to do next, you just give it a high level goal, and it will find its way there.

02:54Typically, the way we've been vibe coding up until now is you might have a goal 10 steps down the path, and you do each step of Claude one by one by one by one. Not anymore. You just give Claude the end goal, and it finds its way there.

03:07So you let it be more autonomous. And then lastly, this is the last tip here Anthropic was giving, is just be more ambitious. You want to now push the limits of what you were doing before.

03:19If you thought something was impossible to do with AI before, go for it now. Be more ambitious, and I will show you how to do that a little bit later in the video as well.

03:29Okay. So let's get into my recommendations on how to get the most out of Claude Fable. I am doing this inside a Claude code in the CLI.

03:37I'm using a little app called Ghosty, which is a nicer version of the terminal to do this. You can feel free to download that for free if you want. I'm hearing from a lot of people Fable is not showing up in the Claw desktop app just yet, so I'd recommend coming in here.

03:52On top of that, I'm hearing from some people Fable is not showing up inside the model picker, So make sure to run this as well slash model Claude dash Fable dash five. Hit enter. Even if it's not in your picker, even if your Claude hasn't updated yet, you can run that, and you will be on the new Claude Fable five.

04:10I recommend going auto mode. So shift tab, go to auto mail mode. It's so smart.

04:15You don't need it to ask you for permissions anymore. I know that makes a lot of people upset. Why?

04:20Why? Why? You're telling people to do dangerous things.

04:22It's smart enough. Do auto mode. It's gonna work.

04:24It's not gonna do anything crazy. From an effort perspective, if you do slash effort, I'd recommend starting off on high. This model absolutely torches tokens.

04:35It does a tremendous amount of thinking. My recommendation Opus was extra high. This, I think you should go with high, and then you can tune it up and down based on the performance you're getting.

04:45So I have it on high auto mode Fable five. So let's talk about how to get the most out of Fable five. What is the best way to put it in these loops?

04:53These loops everyone's talking about. This is the strength of fable. It is so smart.

04:59You don't need to give it little micro steps. You give it a massive goal, and it attacks that massive goal. So you can follow along with me here.

05:06Let's do this. What we're going to do is describe what we want to build. Well, we're not gonna build it just yet.

05:12What we're going to do is use what I call advanced plan mode first. This isn't just a regular plan mode. This is an advanced plan mode where I'm going to have Claude Fable five ask me a bunch of questions to fully understand in-depth what I want to build.

05:30You should be doing this too. Once we have this advanced plan mode built out, then we will give it a goal, which is the big loop it runs in to build this huge idea. So what we're gonna build today is a personal productivity app.

05:44This is everything. This is your to do list, Kanban board, calendar, note taking app, like your Notion on top of that.

05:52This is everything. We're building an entire complex app, and we're gonna build it all in one goal. So, typically, when using, like, Opus, you would build each one of those tools out one by one by one.

06:01Fable, we're doing it all at once. So watch how we do this. So here's the prompt we're gonna use.

06:06I'll put this down below so you can copy and paste it. Feel free to pause right here and grab it so you can do this with me. I wanna build out a personal productivity tool.

06:14I wanted to include to do list, Kanban board calendar, comprehensive note taking app, Pomodoro timer, and even a little game I can play to take a break. Before building anything, please ask me any questions you have on this idea so we can build a fully fleshed out app. I'm gonna hit enter, and it is going to first start asking me any questions it has to develop this idea.

06:36This is not only a good exercise to get the AI up to date on what you wanna build, but also it's a good exercise for you to kind of think through things to fully flesh this out. So where do we wanna build this?

06:48Let's make this a web app. How should your data be stored? Let's make it local only so it's nice and easy.

06:53What how connected should the features be to each other? Definitely deeply integrated. What kind of break game sounds fun?

07:00Let's do a classic arcade game. I like that. And now it fully fleshes out, and we're gonna submit our answers.

07:05Now it's doing more thinking, and now it has even more questions for us. Typical plan mode, just a few questions, then it builds. Here, we're asking to ask as many questions as it needs to get this working.

07:17What visual style do you want? Clean and minimal, playful and colorful, retro terminal.

07:22I kinda like the idea of playful and colorful. Let's do that. How powerful should the note taking be?

07:27Let's do a daily journal style. What should the calendar handle? Definitely task and events.

07:33And anything special for the Pomodoro timer, let's do attach it to tasks as well as stats and streaks. I like that as well.

07:41And then let's hit submit, and we'll submit answers there. Alright. Look at this.

07:45It gave me an entire massive command, uh, to build this out. Let's do this.

07:51We're gonna copy and paste this. It has everything in it. And then what we're going to do is slash goal and then paste that in and hit enter.

08:01And then now what it's going to do is put it in a loop to build out this entire thing. So what's technically gonna happen here is now the agent will go start building this app out, and every once in a while, check back in, okay. Is it accomplishing the goal we gave it?

08:18Is it has it hit the goal in the success criteria in our prompt? This is gonna ensure it does not stop working until the success criteria is met of what's inside the prompt.

08:30And because we did the advanced plan mode, it has a really comprehensive view into what we actually want built. That's the key to doing any of these advanced tasks is the AI no offense to you guys.

08:43I mean, you're really smart because you watch my channel, But the AI at this point, probably smarter, so you should lean on it to do any sort of thinking and planning. Right?

08:53You giving it a command to do something advanced, it probably can command itself to do that advanced thing better. So you wanna say, hey. How would you do this?

09:01What's the best way to do this? Can you plan this out for me? Alright.

09:04Looks like it is all done. It is built, verified, and running and named it Fable. I think maybe, uh, Anthropic might be feeling themselves a little bit because the model's named Fable, and it's now naming all the apps it builds Fable.

09:17Uh, let's take a look at this. Let's see how it's looking. So I'm gonna be honest.

09:20I think the UI is a little ugly, but to be honest, I also answered the question to make it colorful and playful. So I guess it is very colorful and playful. Let's see what we got.

09:30It it looks like it has absolutely everything in it. So let's do a quick task. Edit the fable five video.

09:36Hit enter on that. Boom. It's on your plate.

09:38That's very, very nice looking. Let's check out the calendar. Full calendar.

09:42Let's give it a an event right now. Check video performance. We'll give it tomorrow.

09:48We'll give it a date for Thursday. We'll give it a color, add the event. Boom there.

09:53This is a fully working calendar. Let's see. Is there a week view?

09:56Week view can add the task there and the calendar invites there. Love that. Notes, Full note taking section.

10:02Feeling great about Fable. There.

10:05Boom. Goes into the past entries. That's very nice.

10:08Pomodoro timer. Let's start our Pomodoro timer. Okay.

10:11The focus. Look. It's got streaks.

10:13It's got everything. It's got a full tracker of how many times you use the Pomodoro timer. Let's check out the arcade.

10:19Alright. It is a snake game. That's a good looking snake game.

10:23Makes the snake longer when I eat the strawberries. I don't know if the strawberries little dig at OpenAI there, but there we go. Let's see what happens when we die.

10:31Is it gonna end there? Boom. Adds our score.

10:34Kanban board. There's our task. Can we click and drag it?

10:36It did everything. It built absolutely everything we asked for. Zero bugs.

10:41It just worse works. The tasks are there. Everything's there.

10:45We have the nice dashboard. The key is goals. The key is loops.

10:50If you are new to goals and loops, do not worry. You go into Claude Code. You ask, hey.

10:56I wanna put this in a loop. I wanna make a goal statement. Please tell me what goal to make or what loop we should be using, and it will give it to you.

11:04I'll give you a quick little bit of an advanced workflow as well. If you're using other tools, other project management tools like linear, like you see here, Claude can plug into it. And an interesting loop we can do is we can say something like slash loop, and then every we'll say one hour check to see if there are any new linear tasks you can take.

11:31And now what it will do is every hour go into our linear, our project management tool, and see if there's any coding task it can take and use. That's a very strong way to use loops in here as well. This is how you need to be thinking in loops because these models are so smart and powerful now.

11:49One last interesting point here. By way, if you learned anything at all, make sure to leave a like down below, subscribe, turn notifications. All I do is make amazing videos about AI.

11:58One last thing I wanna talk about real quick is this is only available in subscriptions for the next two weeks on June 22. They take it out of the subscriptions, and you have to pay via API.

12:10It is double the price as Opus. It's actually a fraction of the price of ChadGBT 5.5 Pro, but it is double the price of Opus.

12:18This might be the beginning of the end of subsidized subscriptions.

12:24That's something interesting to watch here. It is the first model a FrontierLab has released that's not available in subscriptions.

12:32So something interesting to know. If you are subscribed at the moment, make sure to test it out. You have a couple weeks to do it, then you'll have to pay via API usage.

12:41That is Claude Fable five. Let me know down below what your experiences have been so far and what you're building with it. I'm very curious to hear.

12:49More videos coming on using it with Hermes and using an Open Claw. Subscribe and turn notifications for that, and I will see you in the next video.

The Hook

The bait, then the rug-pull.

The day Claude Fable 5 shipped, Alex Finn opened his terminal and built a six-feature productivity app in a single command. This breakdown tracks the workflow behind that demo — and the mental model shift that makes it possible.

Frameworks

Named ideas worth stealing.

04:56concept

Advanced Plan Mode

A prompt pattern that asks Claude to interrogate your requirements before writing any plan or code. Forces discovery of edge cases the user has not considered.

Steal forAny complex build where scope creep or missed requirements are the biggest risk.

08:00model