Back to Blog

Claude Fable 5 and Mythos 5: Anthropic's New Models, Explained for Builders

On June 9, 2026, Anthropic released Claude Fable 5 and Claude Mythos 5, its most capable generally available models so far and the first public release of its frontier Mythos class. The headline is software engineering. On SWE-bench Pro, a test of difficult engineering tasks, Anthropic reports Fable 5 scores 80.3%, against 58.6% for OpenAI's GPT-5.5 (Anthropic, via VentureBeat). For teams that build software with AI, that gap is the story. Here is what shipped, what the numbers actually say, and what it changes for builders.

Fable 5 vs Mythos 5: What Is the Difference?

Anthropic built one frontier model and shipped it in two configurations.

Claude Mythos 5 is the full strength model. Anthropic has lifted certain safeguards on it and released it only to approved users, starting with its Project Glasswing cybersecurity partners and a small set of biology researchers.

Claude Fable 5 is that same model made safe for general use. It wraps the Mythos class capability in new classifiers and routing. Most sensitive biology and chemistry requests fall back to Claude Opus 4.8, offensive cybersecurity tasks are blocked, and attempts to distill the model's capabilities into competing systems are refused.

The practical takeaway is simple. When you call Fable 5 through the API, you get frontier capability with guardrails in front of it. The raw model underneath is the same one the approved partners use.

The Benchmarks

Here are Anthropic's reported results, with the comparisons it published:

Two things stand out. First, the SWE-bench Pro lead over GPT-5.5 is wide, not marginal. Second, the FrontierCode Diamond gap, more than double Opus 4.8, suggests the gains are concentrated in long, multi step agentic work rather than single shot answers. Anthropic also calls Fable 5 the best available model for tasks involving vision, citing the ability to pull numbers out of scientific figures and rebuild web app source code from a screenshot.

As always with vendor published benchmarks, treat them as the ceiling and validate on your own workloads. The spread here is large enough to matter.

What It Does in Production

Benchmarks set expectations. Production reports confirm them. Anthropic's launch partners reported the following.

• Stripe said Fable 5 compressed months of engineering into days, completing a migration of a 50 million line Ruby codebase in a single day, work it had estimated at two months for a team.
• Cursor's Michael Truell said it opened up a class of long running problems that were out of reach for earlier models.
• GitHub's Mario Rodriguez said it exceeded previous benchmarks on complex, long running coding tasks.

On the science side, the less restricted Mythos 5 reportedly accelerated protein design by around ten times, with nine of 14 protein targets yielding strong drug design candidates. In blinded comparisons, scientists preferred its molecular biology hypotheses around 80% of the time.

The common thread is duration. These are not faster autocomplete results. They are tasks that run for hours across many steps and hold context the whole way.

Pricing and Availability

Fable 5 and Mythos 5 are both priced at $10 per million input tokens and $50 per million output tokens, with batch processing at $5 and $25. Anthropic notes this is less than half the price of the Claude Mythos Preview, though it remains the most expensive of the major models on the market.

Here is where each model stands at launch.

• Claude Fable 5 is live immediately on the Claude API, consumption based Enterprise plans, Amazon Bedrock, and GitHub Copilot. It is included at no extra cost on Pro, Max, Team, and seat based Enterprise plans from launch through June 22. From June 23 it moves to usage credits.
• Claude Mythos 5 is restricted to Project Glasswing cybersecurity partners and select biology researchers.

Safety and Access

The reason there are two models is safety. Anthropic kept full Mythos 5 access narrow and shipped Fable 5 with new classifiers covering three areas. They block progress on exploitation and offensive cyber tasks, they fall back to Opus 4.8 for most biology and chemistry requests, and they block attempts to extract the model's capabilities for competing models.

External red teaming, Anthropic reports, found zero harmful single turn requests complied with when testers used 30 public jailbreak techniques. For teams in regulated industries, that posture matters. As always, your own evaluations and guardrails sit on top of the model's, not instead of them.

What Fable 5 Changes for Teams Building With AI

We do not read model launches from the sidelines. At Unico Connect, AI is how we build. Our engineers use Claude Code for development, reviews, and tests every day, and we ship AI integrations and agentic systems for clients worldwide. So we read a release like this through one lens: what does it change in the work we are already doing? For Fable 5, the answer concentrates in two places.

The first is large migrations and refactors. The kind of big, mechanical, risky work that used to be quoted in months, things like framework upgrades, language migrations, and dependency overhauls, becomes a supervised exercise measured in days. Stripe's migration is the proof point. The second is longer agentic workflows. Higher reliability on long, multi step tasks raises the ceiling on what an agentic system can own from end to end before it has to hand off to a person.

Vasim Gujrati, our Solutions Architect for AI and Platforms, puts it this way:

"We have built our engineering around AI assisted, long running work for two years, so a model that reliably runs multi hour tasks is not a novelty for us. It is a force multiplier on a workflow we already run. The teams that win the next year will not be the ones with access to Fable 5, because everyone has that. They will be the ones who already know how to wrap a frontier model in guardrails, evaluations, and human checkpoints. That is the part that takes real engineering."

The caution is the one we apply to every agent we build. Capability is not the same as production readiness. A more capable model makes the demo easier. It does not remove the need for guardrails, cost controls, human escalation, and evaluation on your own data. The gap between a working demo and a system you can trust in front of customers is still closed by engineering, not by a model upgrade alone. That is how we approach AI development.

Where Fable 5 Fits Best

A more capable model is not automatically the right model for every job. Based on the work we do for clients, here is where Fable 5 earns its premium price, and where we would reach for something cheaper.

It is a strong fit for:

• Large migrations and modernization. Framework upgrades, language migrations, and legacy refactors are exactly the long, high risk work that used to take months. Stripe's codebase migration shows what becomes possible when one model holds the whole context and runs for hours under supervision.
• Agentic workflows that span many tools. When a task needs an agent to plan, call several systems, recover from errors, and stay coherent across dozens of steps, the reliability gain is worth paying for. This is where most agent projects break, and where the benchmark lead actually shows up.
• Repository wide engineering tasks. Test backfills, large refactors, security and dependency audits, and documentation generation across a big codebase all benefit from a model that can reason over millions of tokens at once.
• Dense knowledge work. Financial analysis, research synthesis, and reasoning over long, complex documents play to its highest scores.

It is probably overkill for:

• Simple chatbots and FAQ assistants. Predefined flows and short answers do not need a frontier model.
• High volume, low complexity classification. Tagging, routing, and sentiment scoring at scale run cheaper and fast enough on a smaller model.
• Short, single step tasks. If the job finishes in one turn, you are paying for autonomy you never use.

The honest rule we give clients is to match the model to the shape of the problem. Fable 5 is built for long, autonomous, high stakes work, so point it at problems that actually have that shape, and keep a cheaper model running everything else.

Frequently Asked Questions

The Bottom Line

Claude Fable 5 moves the frontier on long running coding work and brings Mythos class capability to anyone with an API key, at half the price of the preview. The benchmark leads are real and the early production reports are striking. For teams building software, the practical question is no longer whether the model is capable enough. It is how to put that capability to work safely and reliably. To explore how Unico Connect builds production AI systems and agents on Claude, see our AI development services.