benchr
For seven months, Anthropic's strongest model lived behind a velvet rope. Claude Mythos Preview hunted vulnerabilities inside Project Glasswing while everyone else worked with Opus. The Opus 4.8 launch post promised Mythos-class models "in the coming weeks," and most people read that as marketing weather. It wasn't. On June 9, Anthropic shipped two models from that class at once. The interesting decision is how it split them.
One model, two names
Fable 5 and Mythos 5 share capabilities but differ in access and safeguards. Fable 5 carries safety classifiers. Current API documentation says a blocked request returns HTTP 200 with stop_reason: "refusal"; it does not silently become an Opus 4.8 answer. Applications can configure a server-side, client-side, or manual retry on another Claude model. Mythos 5 omits those classifiers and remains limited to approved Project Glasswing organizations.
Mythos 5 has those safeguards lifted, and you can't have it. It replaces Mythos Preview inside Project Glasswing, deployed with the US government, restricted to vetted organizations. Anthropic calls its cybersecurity capabilities the strongest of any model in the world, which is exactly why it stays gated.
What the numbers say
Anthropic's launch table reports the Mythos-class pair at 80.3% on SWE-bench Pro and lists Opus 4.8 at 69.2%. These are provider-reported results, not independent benchr measurements. The table also reports the higher of Mythos 5 and Fable 5 in each row, says the models remain within 1–3 points, and uses Mythos 5 for starred cybersecurity and biology rows. Treat the figures as a reason to run a representative evaluation, not as a guaranteed improvement on your repository.
The partner reports point the same direction. Stripe says a migration across a 50-million-line codebase compressed from months of engineering into days. Cognition reports the highest FrontierCode score among frontier models, even at medium effort. Hebbia reports the highest score it has measured on its finance benchmark. Vendor-supplied numbers, all of them. But they point one way, and the SWE-bench Pro figure is published with the methodology in view.
Fable 5 vs Opus 4.8 vs Mythos 5
| Spec | Claude Fable 5 | Claude Opus 4.8 | Claude Mythos 5 |
|---|---|---|---|
| Price (in/out per 1M) | $10 / $50 | $5 / $25 | $10 / $50 |
| Context window | 1M tokens | 1M tokens | 1M tokens |
| Max output | 128K | 128K | 128K |
| SWE-bench Pro | 80.3%* | 69.2% | 80.3%* |
| Availability | API, Bedrock, Vertex, Foundry | Everywhere | Glasswing only |
| Restrictions | Classifier can return an explicit refusal; fallback is configurable | Standard | No Fable classifiers; access vetted |
*Anthropic reports the higher of the two models per row; they're within 1–3 points of each other.
Is double the price worth it?
Run the math on a real workload before deciding. A coding agent that burns 2M input and 400K output tokens a day costs about $40 daily on Fable 5 against $20 on Opus 4.8. The cost calculator will do this for your volumes, and the Fable 5 pricing breakdown covers the caching math and the tokenizer wrinkle in detail. The question is whether the capability gap pays for the spread, and the honest answer splits by task.
The best candidates are long-horizon coding, large migrations, and research-style analysis where Anthropic reports the largest gains. Whether those gains survive on your tasks is an evaluation question. Routine chat, drafting, summarization, and high-volume production traffic may fit Sonnet or Haiku at lower cost. One more wrinkle: Fable 5 uses the Opus 4.7 tokenizer, which produces roughly 30% more tokens than pre-4.7 models for the same text. If you're comparing against an older Claude bill, the effective gap can be wider than the sticker price suggests.
The launch inclusions are over
The original June 9–22 inclusion was interrupted by the suspension. After restoring access July 1, Anthropic included Fable 5 for up to 50% of weekly usage limits on eligible paid plans through July 7. Both periods are now historical. Current Claude-plan use requires usage credits; evaluate against a stated credit budget and verify the live plan terms before purchase.