The Free Ride Is Over: Inside Claude Fable 5's Credit Shock and Anthropic’s New ID Checks

Photo by Planet Volumes on Unsplash

Anthropic’s most powerful chatbot (that they marketed as some “apocalyptic doomsday tool” to their own demise) just got a pricey personality change. If you are big on autonomous usage and agentic workflows, your pockets may already know about this before you do.

Since the launch on June 9, 2026, Claude Fable 5 and Claude Mythos 5 have lived through a suspension that saw an Anthropic customer suing the U.S. over loss of Fable AI access, a government export order, a five-day credit deadline, and now a last-minute reprieve that pushed the deadline back again.

In addition to the new pricing system, there is a new identity verification system that requires some users to produce a government-issued ID before they can continue using Claude. This part has obviously raised some eyebrows.

As the subsidy arrangement winds down, subscribers are starting to notice how big the actual bill is.

The free ride, and why it kept getting interrupted

When Fable 5 was launched, it was included for Pro, Max, Team, and seat-based Enterprise plans at no extra cost, but with specific weekly usage limits.

That sounds simple enough, but that arrangement did not prevail for long.

On June 12, the Commerce Department ordered Anthropic to suspend access to Fable 5 and its safety-hardened sibling Mythos 5, but this ban was only meant to apply to foreign nationals anywhere in the world. You can read more about the suspension here.

The suspension continued until June 30 when the export controls were lifted. Fable 5 made its global online comeback on the 1st of July.

Around this time, Anthropic set a short included-access window: up to 50 percent of weekly usage limits through the 7th of July, after which the model would move entirely to metered usage credits.

But as the deadline hit, user backlash over the abrupt switch pushed Anthropic to extend the free window again, this time to the 12th of July at 11:59 PM Pacific.

That makes a total of four access changes in a little over a month.

If you happen to read this piece after July 12, assume that the credits requirement is live unless Anthropic has decided to stick to its new pattern of moving the goalposts again and again.

What Fable 5 actually costs, and how fast it adds up

Photo by engin akyurt on Unsplash

From the minute usage credits kick in, Fable 5 will bill at $10 per million input tokens and $50 per million output tokens. These figures are roughly twice Opus 4.8’s rate and have taken the cup as the steepest pricing Anthropic has introduced for a model outside a limited research preview.

Cached input reads drop to $1 per million, and batch processing cuts both rates in half for work that can wait. This setup makes it a bit more bearable for repetitive, high-context jobs.

Let’s make a comparison using something modest: a planning pass that reads 200,000 tokens of context and writes 40,000 tokens back totals at about $4 on Fable 5. The same job on Sonnet 5's introductory rate costs about $0.80. This is a five times difference that adds up rapidly when you start doing multiple passes a day.

A flat monthly subscription used to translate to predictable costs. But now, a metered credit balance when your autonomous agent does not stop to check in with you means that costs are more unpredictable and you are more likely to haemorrhage money you cannot afford to lose.

The ID checks, and why the timing might look suspicious

This bit of news is separate from the credit switch, but it’s also causing a bit of a buzz in the AI community. Anthropic updated its privacy policy to allow it to request government-issued ID and biometric verification from users, effective July 8.

This setup is run through a third-party vendor (Persona), and it can involve uploading a passport or driver’s license, including a live selfie. Anthropic collects the document image, the printed personal details, and what the policy calls facial geometry templates.

Anthropic has denied any claims that this is in any way tied to the recent ban, and I am inclined to agree. The Persona partnership was first announced back in April, which is months before the export order, as part of a broader age-verification push.

At the time this was written, Anthropic has not validated any theories about links between identity verification and Fable 5 access. The Fable 5 restoration on July 1 traces back to the export-control lift itself, not to any ID-based workaround.

It is best to approach the citizenship-verification theories as speculation because so far, that is what they are.

The takeaway

If you happen to use Fable 5 for anything other than light prompts here and there, I’d advise that you check your usage dashboard at console.anthropic.com before you assume that your plan still covers it. The second you enable credits, make sure you set a hard spending cap.

For routine work, use Sonnet 5 or Opus 4.8 instead. It’s also not a bad idea to watch the calendar and take a peek at the Anthropic website here and there. They have already shifted deadlines once, so it’s not far-fetched to assume that they might do it again.

Confirm the current cutoff before planning around it rather than trusting any single article, including this one.

If a verification prompt pops up on your window asking for your passport, just know that it’s real. It’s highly unlikely to be a phishing attempt from a hacker sitting in some dark room somewhere. That ID verification request is tied to a documented privacy-policy change effective from July 8.

Anthropic, which just had its first profitable quarter is looking for ways to make more money, and usage credits appear to be part of that strategy.