The Reality of Fable 5: Why the Market's Priciest Model Woke Up 'Included' in Your Plan
Your bill changed today
Starting today, July 20, access to Claude Fable 5, the priciest model on the market, has been split in two. If you pay for Anthropic's high-end plans (Max or Team Premium), the model woke up "included" in your subscription, though at half the usage limits. If you are on Pro or Team Standard, you now pay per use at API rates: $10 per million input tokens and $50 per million output tokens, with a one-time $100 farewell credit.
The obvious question: why would a company include its most expensive product "for free"? Anthropic's short answer is accessibility. But the dates tell a fuller story, and it is worth reading because it hits your wallet directly.
Six weeks of roller coaster
A company does not change the access rules of its flagship product five times in six weeks because everything is going according to plan. Something was moving outside.
What moved outside? The sandwich dates
On July 9, OpenAI took GPT-5.6 Sol to general availability: comparable performance, half the price ($5/$30), and included in regular ChatGPT subscriptions, even the $20 Plus plan. On July 16, Moonshot launched Kimi K3 at a third of the price ($3/$15), took the number one spot in frontend code, and announced open weights for July 27.
Coincidence that four days after Kimi's launch, Fable wakes up included in the high-end plans? Maybe. I am not in Anthropic's meetings and I will not invent motives. What I can do is quote how the specialized press read it: The Decoder describes it as a limits cut that pushes Pro users toward API pricing, Rundown AI headlined that Fable "survived the subscription axe", and KuCoin framed it directly "amid the AI price war". The shared reading is churn defense, not philanthropy.
And one piece of context I made in the Kimi article and will repeat: costing 3.3 times more does not mean being 3.3 times worse business. Fable is still number one in overall capability. The question is never which model is best, but which model is best for each of your processes.
What to do with this in your business
Full transparency: I write this as a heavy Claude user in my own operation, and that is exactly why I want to tell it without romanticizing it. My thesis has been the same for two weeks, and this is the third time the market validates it: portfolio, not flag. Providers will keep moving prices, limits and rules; your edge is not guessing the next move, it is running an operation that survives any of them.
If you want to build that map for your company (which model goes in each process and what each one costs you), that is the work I do every day, and soon in person: join the workshop waitlist (held in Spanish) or let's talk directly.
Sources
- The Decoder, Anthropic slashes Claude Fable 5 limits in Max and Team Premium
- TechTimes, Claude Fable 5 Billing Splits Today (July 20, 2026)
- The Rundown AI, Anthropic's Fable survives the subscription axe
- Digital Applied, Claude Fable 5 Pricing: The Usage-Credits Switch
- Forbes, Claude Fable 5 Extends By Five More Days
