Claude Fable 5 and Claude Mythos 5: Anthropic's Most Capable Models Yet
Anthropic has launched Claude Fable 5 and Claude Mythos 5—its most capable models to date. Fable 5 is state-of-the-art on nearly all tested benchmarks, with exceptional performance in software engineering, knowledge work, vision, and scientific research. Mythos 5 is the same underlying model but with certain safeguards lifted for a small group of cyberdefenders and infrastructure providers. Together, they represent a significant leap in what AI can do autonomously over long, complex tasks.
What Makes Fable 5 Different
Fable 5 sets a new bar for autonomous, long-horizon work. The longer and more complex the task, the larger Fable 5's lead over previous models. Stripe used it to perform a codebase-wide migration across a 50-million-line Ruby codebase in a single day—a task that would have taken a whole engineering team over two months by hand. It ships with conservative safeguards for high-risk capability areas. Releasing a model this capable comes with risks: without guardrails, Fable 5's cybersecurity knowledge could be misused. Anthropic launched with safeguards that route certain queries to Claude Opus 4.8 instead. These trigger in less than 5% of sessions on average, though they are intentionally conservative to prioritize safety at launch. Pricing is less than half that of Claude Mythos Preview. Both models are offered at $10 per million input tokens and $50 per million output tokens.Key Capability Highlights
Software engineering. On Cognition's FrontierCode evaluation—which tests whether models can pass difficult coding tasks while meeting production-quality standards—Fable 5 scores highest among frontier models even at medium effort. It's more token-efficient than any previous Claude model. Knowledge work. Fable 5 tops Hebbia's Finance Benchmark for senior-level reasoning, with substantial gains in document-based reasoning, chart and table interpretation, and problem solving. IMC found it aced their trading-analysis evaluations nearly across the board, including factual lookup, root-cause analysis, and expected-value analysis. Vision. Fable 5 is the new state-of-the-art for vision tasks. It can extract precise numbers from detailed scientific figures, rebuild a web app's source code from screenshots alone, and completed Pokémon FireRed using only raw game screenshots—no maps, navigation aids, or extra game-state information. Earlier Claude models required a complex helper harness to do the same. Memory and long-context. Fable 5 stays focused across millions of tokens in long-running tasks and improves its outputs using its own persistent notes. When tested on the deck-building game Slay the Spire, file-based memory improved Fable 5's performance three times more than it improved Opus 4.8's—and Fable reached the game's final act three times more often.Claude Mythos 5: For Cyberdefenders and Researchers
Mythos 5 launches through Project Glasswing in collaboration with the US government. It upgrades Claude Mythos Preview with the strongest cybersecurity capabilities of any model in the world. Anthropic intends to expand access through a broader trusted access program soon. Drug design accelerated by ~10x. Using Mythos 5 with protein design and bioinformatics tools, Anthropic's protein design experts matched or beat skilled human operators without human assistance. Nine of 14 protein targets yielded strong drug design candidates now under investigation. First model to consistently produce novel scientific hypotheses. In blinded head-to-head comparisons against Opus-class models, scientists preferred Mythos 5's molecular biology hypotheses ~80% of the time. One Mythos hypothesis—a novel mechanism for an E. coli protein—was independently corroborated by another research lab. Autonomous genomics research over a week. Mythos 5 assembled single-cell data for millions of cells across 138 animal species, then designed and trained a custom ML model to identify cells performing the same biological role in distantly related organisms. Its trained model outperformed a recent Science-published model—despite being 100 times smaller.Alignment Assessment
Anthropic's automated alignment assessment found Mythos 5's level of misaligned behavior—including deception and cooperation with misuse—to be low and similar to Opus 4.8. Since Fable 5 is the same underlying model, alignment characteristics are expected to be the same. Full details are available in the model's system card.
Source: Anthropic — Claude Fable 5 and Mythos 5 Note (Jun 12, 2026): Anthropic has temporarily suspended access to Claude Fable 5 and Claude Mythos 5 while working to restore availability as soon as possible.