Claude Fable 5: What It Is and How to Get Access Today
If you use Claude for serious work, yesterday's release matters. Anthropic launched Claude Fable 5 on June 9 — making its most powerful model ever available to paying subscribers. This is not a routine version bump. Fable 5 belongs to a class of models Anthropic calls "Mythos," which the company has kept restricted to a small group of enterprise partners until now.
Here is what changed, what the limits are, and how to access it before the free window closes.
What "Mythos-Class" Actually Means
Anthropic organizes its models in tiers. The Opus and Sonnet lines are what most Claude users have worked with. Mythos is the layer above those — the most capable models the company builds, previously available only to select enterprise accounts under restricted terms.
Fable 5 is the first Mythos-class model available to the public.
Anthropic's official announcement describes it as state-of-the-art on nearly all tested benchmarks. The practical translation: Claude is now genuinely competing with GPT-5 and Gemini Ultra at the frontier tier. If you had started to feel like Claude was falling behind on harder tasks, that gap just closed.
What It Does Noticeably Better
Long documents and sustained context. Fable 5 maintains focus across millions of tokens. Anthropic reports three times the performance gains over Opus 4.8 on memory-dependent tasks. If you have worked with Claude on long documents and noticed it losing the thread partway through, this is the improvement most people will feel immediately.
Coding and structured reasoning. The clearest external proof point: Stripe used an early version of Fable 5 to complete a migration of its 50-million-line Ruby codebase in a single day. The same project, using a normal engineering team, would have taken two months. You are probably not migrating a codebase that size, but the underlying capability shows up clearly on any complex, multi-step programming task. Our guide to the best AI chatbots for coding covers how to structure requests to get the most out of Claude for development work.
Visual analysis. Fable 5 can look at a screenshot of a web application and rebuild it from scratch. It can extract precise numbers from charts, graphs, and scientific figures. If you regularly send images to Claude and want more than a surface-level description, the results are substantially better.
Knowledge work. On Hebbia's Finance Benchmark — a test designed to evaluate senior-level financial reasoning — Fable 5 scored higher than any other tested model. Document-heavy analysis, table interpretation, and tasks that require combining information across many sources are where the model makes the clearest jump from what came before.
The Safety Limits (And When You'll Hit Them)
Fable 5 includes hard classifiers that block three categories of requests: cybersecurity exploitation tasks, biological and chemical dual-use content, and attempts to extract the model's training data. When one of these triggers, the model falls back to Claude Opus 4.8 for that response rather than refusing entirely.
This happens in less than 5% of sessions. For almost everything you would normally use Claude for — writing, research, analysis, coding, brainstorming — you will never encounter these limits.
One thing worth noting if you have an enterprise account: Anthropic now requires 30-day traffic retention for all Fable 5 users, including accounts that had previous zero-retention agreements. Anthropic says this is necessary to defend against novel jailbreak attempts. If you had a zero-retention contract, check with your account manager before upgrading to Fable 5 at scale.
How to Get Access Right Now
The access window makes the timing important.
Through June 22: Fable 5 is included at no additional cost in Claude Pro, Max, Team, and Enterprise plans. If you are a paid subscriber, you can switch to Fable 5 now. Anthropic is rolling it out across subscription tiers through June 22 — if you do not see it yet, you should within the next few days.
After June 22: Access will require credits. The API rate is $10 per million input tokens and $50 per million output tokens — double the cost of Opus 4.8, but less than half the price of the Mythos Preview tier that preceded it. Anthropic has not published credit pricing for subscription users after the free window, so expect an announcement before the 22nd.
Free plan users: Fable 5 is not on the free tier. You will need a paid subscription or API access to try it.
For comparison: Claude Pro is $20 per month, in the same range as ChatGPT Plus and above Meta's $7.99 Meta One Plus plan (which we reviewed when it launched last month). The free Fable 5 access window makes the next two weeks a good time to test whether Pro is worth it for your workflow. You can check Claude's capabilities and current plans on its chatbot.gallery profile.
What About Mythos 5?
You may have seen Claude Mythos 5 mentioned alongside Fable 5. It is the same underlying model — identical weights, identical capabilities — but with the cybersecurity classifiers removed.
Mythos 5 is not available to the general public. It is being deployed through Project Glasswing to cyberdefenders, infrastructure security teams, and eventually a broader group of vetted researchers. If you are not working in that area, you do not need to think about it. Fable 5 is what you want. TechCrunch's coverage has more context on why Anthropic chose to release both models at this particular moment.
What to Do Now
If you are a paid subscriber: switch to Fable 5 and try it on a task you know well. Complex documents, multi-step coding problems, and anything requiring sustained reasoning are the clearest ways to feel the difference from Opus 4.8.
If you are on the free tier: this is a reasonable time to test a one-month paid subscription. The free Fable 5 window is a real advantage while it lasts.
If you use the API: the Anthropic documentation covers updated model IDs and migration notes.
The About.chat Weekly newsletter covers AI model releases, pricing changes, and what matters for everyday users — free, every week. Subscribe here.
Stay in the loop
Get the best chatbot news, reviews, and discoveries — weekly.