Claude Fable 5 Is Now Available on WidelAI
Claude Fable 5, Anthropic's Mythos-class flagship, is back online and live on WidelAI. Here's what it brings, how it's priced in credits, and how to start using it in under a minute.
WidelAI Team
Building the future of AI accessibility
Claude Fable 5 Is Now Available on WidelAI
Anthropic's most capable model is back online — and it is live on WidelAI today. Claude Fable 5 is in your model picker right now under the tag claude-fable-5, with no waitlist and no separate Anthropic subscription. After a dramatic month that saw Fable 5 launch, get pulled worldwide under a US export directive, and then return, you can finally put the frontier's most talked-about model to work. This post covers what Fable 5 is, how it got here, how it is priced on WidelAI, and how to start using it in the next minute.
Claude Fable 5 is built by Anthropic. Read the official announcement on Anthropic's redeployment note and the Claude model documentation.
What Is Claude Fable 5?
Claude Fable 5 is the first publicly released model in Anthropic's new Mythos-class tier — a step above the Opus line and, by Anthropic's own description, the most capable model it has ever made generally available. It is built for the hardest category of AI work: long-horizon, autonomous reasoning and agentic tasks that previous models could not sustain without losing the thread.
The headline capabilities:
- 1 million token context window, so entire codebases, document sets, or multi-hour sessions fit in a single run
- Up to 128K output tokens per request for large, structured generations
- Adaptive thinking by default — no separate "extended thinking" toggle to manage
- Multimodal input with strong vision, plus persistent focus across long-running tasks
- Purpose-built for multi-day autonomous work, large codebase migrations, and complex knowledge synthesis
On benchmarks, Fable 5 sits at the top of agentic-coding evaluations:
| Benchmark | Claude Fable 5 | Claude Opus 4.8 | GPT-5.5 |
|---|---|---|---|
| SWE-Bench Pro | 80.3% | 69.2% | 58.6% |
| FrontierCode | 29.3% | 13.4% | — |
Those are not marginal gains. An 11-point jump on SWE-Bench Pro over the previous Opus generation is the kind of step change that shows up in real agentic workflows — fewer failed runs, longer autonomous stretches, and less babysitting on complex tasks.
The Road Back: From Export Ban to Global Return
If you followed the story, you know Fable 5's launch was anything but ordinary. Here is the short version:
- June 9, 2026 — Anthropic ships Claude Fable 5, its first public Mythos-class model, alongside a more tightly controlled sibling, Claude Mythos 5.
- ~June 12, 2026 — Three days after launch, the US Department of Commerce issues an export-control directive citing national security, requiring Anthropic to restrict access for foreign nationals. Rather than build nationality gating overnight, Anthropic suspends the models for everyone.
- June 30, 2026 — After weeks of negotiation, the Commerce Department lifts the export controls on Fable 5 and Mythos 5.
- July 1, 2026 — Anthropic begins restoring global access across its platforms.
We covered the original disruption — and what it means for building a resilient AI strategy — in The Future of Fable 5 and the Rise of GLM-5.2 Open Weights. Now that access is restored, WidelAI has added Fable 5 so you can use it the moment it is available again, without renegotiating your own access or standing up a new billing relationship.
Fable 5 Pricing on WidelAI
WidelAI uses a transparent credit system: 1 credit equals $0.005 of real API cost, and input and output tokens are priced separately because generating output costs far more than reading input. Fable 5 is Anthropic's most powerful — and most expensive — generally available model, and its rates reflect that:
| Model | Input (credits / 1K tokens) | Output (credits / 1K tokens) |
|---|---|---|
| Claude Fable 5 | 2.00 | 10.00 |
| Claude Opus 4.8 | 1.20 | 6.00 |
| Claude Sonnet 5 | 0.60 | 3.00 |
These rates come directly from Anthropic's published list price of $10 per million input tokens and $50 per million output tokens — roughly double the input and triple the output cost of Opus 4.8. Fable 5 is a model you reach for when the task genuinely warrants frontier capability, not for routine chat.
The upside: you do not pay Anthropic separately, and you do not manage Fable 5's access rules yourself. One WidelAI plan puts Fable 5 alongside top models from OpenAI and Google, with credits that work across all of them. See the full breakdown on our pricing transparency page.
What Fable 5 Is Good At
Fable 5 earns its premium on tasks where failure is expensive and depth matters:
- Large-scale codebase work — migrations, multi-file refactors, and debugging across sprawling repositories that need the full 1M context
- Long-running agentic sessions — chains of steps that would cause earlier models to drift or lose state
- Deep research and synthesis — pulling signal from large document sets in a single pass
- Complex reasoning — architecture decisions, tricky analysis, and problems where getting it right the first time is worth the extra cost
For everyday chat, drafting, and high-volume tasks, a Sonnet- or Flash-class model is the smarter economic choice. Fable 5 is the tool you escalate to.
How to Use Claude Fable 5 on WidelAI
Getting started takes about a minute:
- Sign in to your WidelAI account, or create one if you are new.
- Open the chat and click the model selector.
- Choose Claude Fable 5 from the Anthropic group, or paste the model tag
claude-fable-5if you are calling it programmatically. - Send your prompt. Adaptive thinking runs by default, so you get deep reasoning without extra configuration.
Fable 5 is available on the Pro plan, right next to Claude Opus 4.8, Claude Sonnet 5, GPT-5.5, and Gemini 3.5 Flash. Browse the full catalog on our models page.
Fable 5 vs the Rest of the Catalog
A quick guide to when to reach for Fable 5 versus its neighbors:
- Choose Claude Fable 5 for the most demanding agentic coding, long-context, and deep-reasoning tasks where frontier quality justifies the cost.
- Choose Claude Opus 4.8 for heavy reasoning and large agentic runs where you want flagship quality at a lower rate than Fable 5.
- Choose Claude Sonnet 5 for the majority of coding, writing, and analysis work that needs strong quality without flagship pricing.
- Choose a Flash or mini model (like Gemini 3.5 Flash or GPT-5 mini) for ultra-high-volume, latency-sensitive tasks where cost per call is the priority.
Because everything lives in one account, you can route easy work to cheaper models and escalate to Fable 5 only when a task demands it — all from the same credit balance.
Access That Can Change Overnight — Why One Platform Helps
Fable 5's month offline is a reminder that access to a single frontier model can vanish for reasons that have nothing to do with your product — a policy change, a pricing decision, or a directive. The teams that weathered June 2026 best were the ones that were not wired to exactly one model.
That is the core value of a multi-model platform. With WidelAI, if any single model becomes unavailable or too expensive for a workload, you switch to another without re-architecting anything. When Fable 5 went dark, multi-model users kept shipping on Opus 4.8 and GPT-5.5. Now that Fable 5 is back, they get it instantly. You stay flexible either way.
Get Started Today
Claude Fable 5 is live on WidelAI right now. Whether you are running large agentic coding sessions, migrating a codebase, or tackling research that needs the full million-token context, it is the most capable model you can reach today.
Already a member? Open the chat, select Claude Fable 5, and try it on your next hard problem.