Anthropic's newest models aren't just quicker and cheaper. In prerelease testing, they rebuilt production software, tackled open questions in physics, and even built an operating system from an empty folder. Here's what Claude Fable 5.2 and Mythos 5.2 can do, how they differ, and who gets access.
Anthropic has upgraded Claude Fable and Claude Mythos, models designed to tackle the most daunting technical challenges. Claude Fable 5.2 and Claude Mythos 5.2 show improvements in coding, reasoning, and vision. They generate outputs up to 30% faster than Claude Fable 5.1 and Claude Mythos 5.1, and they cost less: $8 per million input tokens and $40 per million output tokens.
During prerelease testing, the models helped Anthropic's engineers complete otherwise-infeasible software migrations, design three working electronic devices from scratch, solve two open problems in the physics of black holes, and make substantial progress on robotics evaluations. They also formalized the proof of Fermat's Last Theorem in 30 hours.
How Fable and Mythos relate
Claude Fable and Claude Mythos are the same underlying model. Fable has strong cybersecurity and biology safeguards to guard against misuse, which allows Anthropic to make it available to all users. Organizations pursuing beneficial applications in cybersecurity or biomedical research can apply for access to Mythos through the recently expanded Life Sciences and Cyber Verification Programs.
Testers found that both models communicate more clearly than past Fable and Mythos models. Both also come with an optional fast mode, which speeds up token generation by up to 2.5 times, at $16 per million input tokens and $80 per million output tokens.
Mode
Input (per million tokens)
Output (per million tokens)
Standard
$8
$40
Fast mode
$16
$80
Safety and alignment
Anthropic ran extensive prerelease safety testing on the underlying model. This included novel evaluations that give deeper insight into specific concerning behaviors like sandbox escapes. Anthropic believes high-stakes risks from misalignment remain low, and the model performed as well as, or better than, recent models on most alignment measures. Fable 5.2's cyber and biology safeguards are as robust as those of Fable 5.1 but produce fewer unnecessary blocks. Anthropic has also implemented its strongest-ever protections against distillation attacks. Full details are in the Claude Fable 5.2 and Mythos 5.2 System Card.
When to use them
Fable and Mythos are best for the most ambitious, open-ended technical work. For most coding and knowledge tasks, Anthropic recommends starting with Claude Opus 5.5, its strongest daily driver. Users can then move up to Fable or Mythos, or down to Claude Sonnet 5.5 or Claude Haiku 5.5, depending on scope and complexity. Fable 5.2 can also serve as an Advisor, where Haiku, Sonnet, or Opus models consult it on harder problems. It can act as an Orchestrator too, delegating simpler tasks to smaller models.
Benchmark performance
Anthropic says Fable 5.2 and Mythos 5.2 are a step above Opus 5.5 in almost all the capabilities evaluated. Scores for Fable 5.2, Fable 5.1, and Opus 5.5 are with safeguards on unless noted, and all scores show each model's highest-available effort unless noted.
FrontierCode 2.0: Fable 5.2 scores 49.1% at Xhigh effort, against 32.4% for Fable 5.1, 45.5% for Opus 5.5, 42.8% for GPT-6 Astra, and 41.1% for Gemini 4 Carbon at Medium effort.
FrontierSWE v2: 69.9%, against 56.3% for Fable 5.1, 62.3% for Opus 5.5, 65.5% for GPT-6 Astra, and 61.2% for Gemini 4 Carbon.
AutomationBench: 51.4%, against 31.4% for Fable 5.1 and 42.5% for Opus 5.5. Gemini 4 Carbon scores higher at 53.9%.
Terminal-Bench-Science 0.1: 74.6%, or 75.3% for Mythos 5.2, against 52.6% for Fable 5.1, 58.7% for Opus 5.5, 68.1% for GPT-6 Astra, and 64.8% for Gemini 4 Carbon.
CyScenarioBench: Mythos 5.2 scores 76.3%, against 61.7% for Mythos 5.1.
CVE Rediscovery: Mythos 5.2 scores 89.6%, against 78.3% for Mythos 5.1 and 81.3% for Opus 5.5 without safeguards.
De novo binder design (24 hr): Mythos 5.2 scores 85.1%, against 79.1% for Mythos 5.1 and 82.6% for Opus 5.5 without safeguards.
WANDR (relaxed score): 82.1%, against 68.7% for Fable 5.1 and 72.3% for Opus 5.5.
OSWorld 2.1 (offline set, partial): 89.3%, against 83.5% for Fable 5.1, 87.2% for Opus 5.5, 79.5% for GPT-6 Astra, and 81.5% for Gemini 4 Carbon.
RoboDojo Sim: 64.9%, against 27.6% for Fable 5.1, 39.4% for Opus 5.5, and 29.8% for GPT-6 Astra.
Remote Labor Index: 41.3% at High effort, against 17.9% for Fable 5.1, 21.3% for Opus 5.5, and 20.8% for GPT-6 Astra.
Humanity's Last Exam (with tools): 74.2%, against 65.6% for Fable 5.1, 67.7% for Opus 5.5, and 57.2% for GPT-6 Astra.
Chartography (no tools): 73.4%, against 44.8% for Fable 5.1, 64.4% for Opus 5.5, 71.0% for GPT-6 Astra, and 72.5% for Gemini 4 Carbon.
Effort and efficiency
On Medium effort, the default in Claude apps and Claude Code, Fable 5.2 can accomplish challenging tasks with fewer tokens than Fable 5.1. Anthropic's early results show gains in cost per task as well:
FrontierCode: The benchmark measures whether a model's code changes would be merged by the maintainers of open-source repositories. At High effort, Fable 5.2 scores about the same as Opus 5.5's best score (45.3% vs. 45.5%) for roughly half the cost per task. It also beats GPT-6 Astra's best score (42.8%) at about the same cost. At Xhigh effort, its 49.1% is the highest score of any released model Anthropic has tested.
WANDR: Perplexity's benchmark measures agents on large data-collection tasks. At Medium effort, Fable 5.2 matches Opus 5.5's top score for about 40% of the cost per task.
BenchCAD 2.1: The benchmark measures hardware engineering, scored 0–100 with partial credit. At Max effort, Fable 5.2 scores 83.5, ahead of Opus 5.5 (65.3) and Fable 5.1 (44.4).
Humanity's Last Exam: At Low effort, Fable 5.2 matches Opus 5.5's top score for less than one-tenth of the cost per task.
Chartography: At Medium effort, Fable 5.2 scores 66.4%, ahead of Opus 5.5 at Max effort (64.4%) for less than one-tenth of the cost per task.
Coding
Over 90% of the changes merged into Anthropic's core codebase are now written by Claude, with human guidance. Anthropic used Fable 5.2 to fix stubborn bugs in minutes and optimize performance across all of its services. It also used the model to complete high-value software projects that would otherwise have been too slow or too costly to attempt.
A team of agents running an earlier version of Fable 5.2 migrated a critical production service from Python to Rust. The service is now more efficient, more reliable, and starts about 26 times faster.
In one experiment, an Anthropic engineer gave Fable 5.2 the task of designing a prototype operating system. Starting from an empty folder, the model coordinated over 1,000 subagents, working around the clock for eight days writing, testing, and reviewing code. It produced a working operating system that boots on a laptop. The system has its own core that runs the machine, drivers for the disk, keyboard, and network, permanent file storage, windows, a login screen, an installer, and built-in Claude functionality.
Cybersecurity
Claude Mythos 5.2 is Anthropic's most capable model for cybersecurity engineering. It catches more vulnerabilities than previous models and is better at reproducing and fixing them. Fable 5.2's safeguards will block most cybersecurity-related tasks.
A security company found that one in six confirmed bugs found by Mythos 5.2 were missed by all of the other frontier models it tested.
On the internal CVE Rediscovery evaluation, Mythos 5.2 found 10.2% more vulnerabilities than Claude Opus 5.5 and 14.4% more than Mythos 5.1.
One tester found that previous models had overstated the seriousness of about 1 in 20 findings, whereas Mythos 5.2 had not overstated any.
CVE Rediscovery assesses how many published CVEs (Common Vulnerabilities and Exposures) a model can independently find when given a snapshot of the project before a CVE was fixed. In Anthropic's October 2026 run against 1,829 CVEs, Mythos 5.2 outperformed Mythos 5.1 and Opus 5.5 at every effort level.
Claude Enterprise customers, even those outside the Cyber Verification Program, can access Claude Security. It scans codebases for vulnerabilities using Anthropic's most capable models, which now include Mythos 5.2. A recent update helps it do the following:
watch repositories as code is updated
track findings to ensure they are fixed
learn from additional context
connect to a user's tools through an API and MCP
Science and engineering
Many tasks in science, engineering, and mathematics increasingly depend on coding, from analyzing data and modeling molecules to designing parts and controlling instruments. Anthropic says Fable 5.2 and Mythos 5.2 have helped researchers make progress on difficult problems in these areas.
In theoretical physics, Fable 5.2 solved two open problems concerning black holes and the effort to unite relativity and quantum mechanics. Both results rely on holographic duality, the idea that certain universes containing gravity can be precisely described by a quantum theory containing no gravity at all.
Black hole formation: Physicists showed in 1983 that as these universes heat up, black holes form abruptly, like water boiling. Their method worked only in large universes. A calculation for very small universes, where only the quantum description can be used, was laid out in 2003 but was too laborious to complete. Fable 5.2, working largely autonomously, completed it and found the transition is abrupt there, too.
Quantum gravity theories: Fable 5.2 found the exact limit of a technique physicists have used for more than 15 years to rule out candidate theories of quantum gravity in a simplified universe.
Key Features at a Glance
Speed: Output up to 30% faster than Fable 5.1 and Mythos 5.1, so long agentic jobs finish sooner.
Price: $8 per million input tokens and $40 per million output tokens, lower than the previous models.
Fast mode: Up to 2.5 times faster token generation at $16 per million input and $80 per million output tokens, for when speed matters more than cost.
Efficiency: On FrontierCode, High effort roughly matches Opus 5.5's best score at about half the cost per task.
Advisor and Orchestrator roles: Smaller Claude models can consult Fable 5.2 on hard problems, or it can delegate simpler t
Anthropic’s Claude Fable 5.2 is an AI model whose launch materials emphasize agentic coding and broader reasoning capabilities. The most directly relevant mainstream benchmark results are therefore coding and reasoning tests; the published figures below come from Anthropic’s benchmark table, with competitor scores included where reported. Results use maximum effort averaged over five trials, except FrontierCode 2.0, which used each model’s highest reported effort setting, 41.3%High effort.
