🔑 Fable 5 Is Gone After Today

Reading time: 5 minutes

Greetings from above,

Why did the AI user spend his last Fable 5 day building a demo app? Because he wanted to impress people who were also using demo apps.

I almost made this mistake myself. I had a full list ready — websites, content batches, a couple of automations — basically a to-do list that Opus 4.8 could've knocked out on a Tuesday morning for free. Then I sat with one question long enough to change everything: can a cheaper model redo this tomorrow? The answer was yes for everything on my list. So I threw the list out and ran five moves that actually leave something behind.

Today, we'll talk about:

Let's get into it.

The AI tool you tried last quarter waited for a prompt, hallucinated a number, then asked if you'd like a summary.

Viktor opened a PR at 2am, rebased it against main, ran your test suite, and posted a note in #eng: "Two flaky tests in payments service, both pre-existing. Recommended merging after fixing them." Then drafted the customer reply for the support ticket the bug created.

That's 619K autonomous actions per day across 20,000+ teams. Not chat replies. Real work shipped to GitHub, Stripe, Linear, Notion, and 3,000+ other tools, from inside Slack and Microsoft Teams.

You don't supervise him any more than you supervise a senior engineer.

SOC 2 certified. Your data never trains models.

"It's what you probably originally thought AI was going to be when you first heard of it in sci-fi movies." Tyler, CEO.

Get Started for Free

Ok, so today — July 7, 2026 — is the last day Fable 5 is included in standard Claude plans at no extra cost.

Starting tomorrow, it moves to usage credits billed at $10 per million input tokens and $50 per million output tokens. That's exactly double the rate of Opus 4.8, which is, by the way, still sitting inside your normal plan limits doing just fine. For most people on a Pro or Max plan, the practical math is actually pretty simple: Fable 5 is basically gone after today unless you specifically enable credits and pay the overage.

So the real question isn't "what should I build today." It's "what can only Fable 5 build — that a cheaper model can then run forever after."

That's a very different question. And the answer is what the next five moves are all about.

Before any move, run this filter on every task you're considering.

Can a cheaper model redo this tomorrow?

A website? Yes. A batch of posts? Yes. A demo app? Absolutely yes. Opus rebuilds all of that next week for nothing. Spending your last Fable hours on that kind of work is, more or less, like hiring a surgeon to take your blood pressure.

What a cheaper model actually can't redo tomorrow is anything that requires Fable-level judgment to create — but only ordinary intelligence to use afterward. A written-down standard. A reasoned-through roadmap. A distilled knowledge vault. An installed skill that fires on its own. These things keep their full value after the model that built them is out of reach.

So the strategy for today is extraction, not conversation.

Move 1 — Plant Fable's judgment in your workspace

The highest-value thing a frontier model leaves behind is a standard. An answer helps you once. A standard upgrades every answer that comes after it.

Your CLAUDE.md, your skills files, your memory setup — this is the layer every future model reads before it touches your work. Today, Fable writes that layer at a level Opus can follow but could never author. Run this in every project you care about:

"Read this entire project and how I work in it. Then rewrite my CLAUDE.md as the operating manual a less capable model would need to work here at your level: the conventions I follow and the ones you'd add, the mistakes a weaker model will make in this codebase with the rule that prevents each one, the quality bar per deliverable written as checkable criteria not adjectives, and what to do when uncertain with the exact escalation rules. Then propose the 3 skills that would save me the most hours, and write them in full."

The criteria line is basically the whole point. A cheaper model can't invent a quality bar — but it applies a written one just fine.

Move 2 — Run the consultant audit

Fable's verified edge is judgment on hard, messy problems. On the hardest coding benchmark tier it scores more than double the next model, and the gap widens as difficulty rises. So give it the hardest messy problem you own: your business.

Open a session with access to your projects, your numbers, whatever context you can feed it, and run:

"Act as the consultant I can't afford. Audit everything: projects, offers, workflows, pricing, where my time goes. Deliver a roadmap I can execute with a less capable model: ranked moves with highest expected return first, per move the exact steps and what done looks like, and the three things I should stop doing with the reasoning written out in full."

The deliverable rule carries the value. The reasoning gets written down today, while the model that can produce it is still flat-rate. Tomorrow, Opus doesn't need to be brilliant — it just needs to follow a brilliant document.

Move 3 — Build the second brain

Research is where the extraction runs deepest. Long, multi-step synthesis is Fable's widest measured lead over every other model. So spend a slice of today on volume: deep research runs on your niche, your competitors, your customers' problems, the methods you keep meaning to study properly.

Then mine every run into an Obsidian vault, one insight per note, each note linking to related ones. A hundred linked one-insight notes get retrieved and reused. A 40-page report gets stored and forgotten. Don't summarize into long documents — atomize. The vault becomes the context every future session reads from.

Move 4 — Fire the goals

Fable's signature capability is holding one job for hours without losing the thread. That endurance is exactly what stops being flat-rate tomorrow — so today it goes to work unattended.

In Claude Code, use /goal to set a finish line rather than a prompt. You describe what done looks like, and the model keeps working turn after turn while a second, smaller model checks the condition after every turn and stops the run only when it's met.

Two rules make this safe rather than expensive. First, demand pasted proof in the finish line — the judge model reads only the conversation, so ask for the green run pasted, never just promised. Second, cap every run — turns or wall-clock time, written into the condition. One unattended loop without a cap can run a very large bill by morning. Pick the two or three goals with the most locked-up value, not ten.

Move 5 — Install the recorder (do this first if you only have an hour)

Every time Fable cracks a hard problem today, its approach evaporates when the session ends. This move installs a recorder before that can happen.

Create the file .claude/skills/extract-approach/SKILL.md and wire it into your CLAUDE.md like this:

"After every non-trivial solved problem, run the extract-approach skill before moving on. A solution without its learnings note is unfinished work."

Now spend the rest of the day working Fable hard on your real backlog — the gnarly bug, the architecture decision you've been circling for weeks. Every solve leaves a note behind. The notes are the distillate: Fable's reasoning sitting in your repo, readable by every model that comes after.

This is the compounding move. It converts all your remaining Fable time into permanent assets automatically.

If time is short, run the moves in this order: 5 first, then 4 (it runs alone while you work), then 1, then 2, then 3.

What you learned today:

The wrong way to spend today is building websites, demo apps, and content batches — a cheaper model rebuilds all of that next week for nothing, and Fable's real edge is the judgment it takes to write a standard, not the speed it takes to write a post

The five moves work because they create outputs that require frontier-level intelligence to produce but only ordinary intelligence to use — the value doesn't disappear when the model does, it just sits in your files waiting for the next model to run on it

This situation will repeat — a frontier window opens, gets re-priced, sometimes goes dark — and the only thing that carries forward is what you extracted while you had access

The teacher is leaving today. What you pull out before the door closes is yours to keep. Every model that comes after Fable will be smarter, cheaper, and more capable — but it won't know your business, your codebase, or your standards unless Fable wrote them down first.

That's what today is actually for. Go run the five moves.

And as always, thanks for being a part of my lovely community,

Keep learning,

🔑 Robert from God of Prompt