Starting today, June 23, 2026, Claude Fable 5 is no longer part of your Claude Pro, Max, Team, or Enterprise subscription. If you want to keep using it, you need usage credits billed at the API rate. If you take no action, your subscription keeps working — you just won’t have Fable 5 anymore.

Here is exactly what changed, what it costs, and what to do about it.

What Changed and Why

Anthropic gave paid subscribers a complimentary window to use Fable 5 from June 9 through June 22. The intent was a 13-day trial before any charges began. What actually happened was messier.

On June 12, a US government export control directive forced Anthropic to suspend Fable 5 access globally. The model was offline until approximately June 18. That means subscribers who wanted to try Fable 5 had roughly four to five days of actual access during the complimentary window — not 13.

Anthropic is moving forward with the June 23 cutover regardless. The model is now available via usage credits only, billed at API rates.

The New Pricing

Fable 5 is now the most expensive model Anthropic ships:

Model	Input (per M tokens)	Output (per M tokens)
Claude Fable 5	$10	$50
Claude Opus 4.8	$5	$25
Claude Sonnet 4.6	$3	$15

That is double the cost of Opus 4.8, which remains the flagship included in all standard subscription tiers. For most professional use — analysis, drafting, research, code review — Opus 4.8 is extremely capable. The gap between Fable 5 and Opus 4.8 is meaningful for edge-of-frontier tasks like complex multi-step reasoning and long-context work, but for everyday professional use, most people will not notice the difference.

What Happens if You Do Nothing

If you have a Claude Pro, Max, Team, or Enterprise subscription and you do not purchase usage credits:

• Your subscription continues working normally
• Claude Opus 4.8, Sonnet 4.6, and other standard models remain available without any change
• Fable 5 still appears in your model picker but selecting it will either prompt you to add credits or fall back to Opus 4.8, depending on your plan settings

Your subscription is not broken. This is an add-on change, not a removal of existing functionality.

What Happens if You Add Credits

Usage credits are billed separately at API list prices. You purchase them outside your subscription, they do not roll over at month end, and once you run out, Fable 5 use stops until you add more.

For teams running occasional Fable 5 sessions — exploratory analysis, capability testing, high-stakes document work — the math is manageable. A $50 credit pool gets you through roughly 5 million input tokens or 1 million output tokens on Fable 5, which is several days of heavy professional use.

For automated workflows or high-volume processing, Fable 5 is now a deliberate line item in your AI budget. It should not be running in the background on tasks where Opus 4.8 would do fine.

Anthropic’s Stated Position

Anthropic has been direct that this is a capacity decision, not a permanent pricing tier. Their stated intent is to restore Fable 5 as a standard part of subscription plans once infrastructure allows, and to communicate changes in advance.

Given the pattern of the past month — government suspension, limited rollout, metered access — “once capacity allows” is doing a lot of work in that sentence. Plan as if this pricing is the baseline for the foreseeable future. If it changes in your favor, that is a bonus.

The Larger Trend This Fits Into

Fable 5 leaving the subscription bundle is the latest step in a shift that has been building since early 2026. Anthropic separated programmatic usage from subscription limits on June 15. GitHub Copilot moved to usage-based billing on June 1. OpenAI has been iterating toward consumption-based enterprise plans for months.

The era of flat-rate AI is not ending — basic access and capable models are still included in subscriptions across every major provider. But frontier access is being metered. The models at the cutting edge are becoming line items that teams need to deliberately budget for, not ambient tools that cost the same whether you use them occasionally or constantly.

This is probably the right direction. Frontier models require enormous compute, and subsidizing heavy usage through flat subscriptions was always economically fragile. The question for business teams is whether they are treating AI usage as infrastructure spending — budgeted, tracked, and optimized — or still thinking of it as a software subscription.

What This Means for Business

For business owners and teams using Claude professionally, the Fable 5 billing change is less about the model itself and more about the shift in how AI costs work.

If your team has been treating Claude as a flat monthly cost, now is a good time to actually audit what you use it for, how often, and whether the models you need justify the spend. Opus 4.8 is the right default for the vast majority of professional tasks. Fable 5 is worth the premium for specific, high-stakes work where reasoning depth genuinely matters.

The teams that will manage AI costs well are the ones that build usage awareness into their operations — knowing which tasks go to which model and why, rather than defaulting to the frontier model on everything because it was included.

If you are thinking through how to structure your team’s AI stack without overspending, that is exactly the kind of decision-making that EDNA’s Omni Advisory service is built for.