Anthropic releases Claude Fable 5.2 and Mythos 5.2 at $8/$40 per 1M tokens, with an optional fast mode

Anthropic has released Claude Fable 5.2 and Claude Mythos 5.2 at $8 per million input tokens and $40 per million output tokens, which Anthropic says is lower than the price of Fable 5.1 and Mythos 5.1. The company says the models generate output up to 30% faster than their predecessors.

Anthropic describes the two as "best for the most ambitious, open-ended technical work."

Same model, different safeguards

According to Anthropic, Fable and Mythos are the same underlying model. Fable carries strong cybersecurity and biology safeguards, which lets Anthropic offer it to all users. Organizations doing beneficial work in cybersecurity or biomedical research can apply for Mythos access through Anthropic's expanded Life Sciences and Cyber Verification Programs.

Anthropic claims Fable 5.2's cyber and biology safeguards are as robust as Fable 5.1's but produce fewer unnecessary blocks. It also says it has implemented its strongest-ever protections against distillation attacks.

Pricing and fast mode

• Standard: $8 per 1M input tokens, $40 per 1M output tokens
• Fast mode (optional): $16 per 1M input tokens, $80 per 1M output tokens, with token generation up to 2.5 times faster

Context window and parameter count are not stated in the announcement.

Benchmarks

Anthropic reports that Fable 5.2 and Mythos 5.2 are a step above Claude Opus 5.5 in almost all the capabilities it evaluated. Scores for Fable 5.2, Fable 5.1 and Opus 5.5 are with safeguards on unless noted, and each model is shown at its highest available effort unless noted. Methodology is in the system card.

• FrontierCode 2.0 (agentic coding): Fable 5.2 scores 49.1% at Xhigh effort, which Anthropic says is the highest of any released model it has tested. At High effort it scores 45.3%, against 45.5% for Opus 5.5's best score, at roughly half the cost per task. GPT-6 Astra's best is 42.8%.
• BenchCAD 2.1 (hardware engineering, scored 0 to 100): at Max effort Fable 5.2 scores 83.5, ahead of Opus 5.5 (65.3) and Fable 5.1 (44.4).
• Chartography (no tools): 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, according to Anthropic.
• Humanity's Last Exam (with tools): at Low effort Fable 5.2 matches Opus 5.5's top score for less than one-tenth of the cost per task, per Anthropic.
• WANDR (Perplexity's agentic data-collection benchmark): at Medium effort Fable 5.2 matches Opus 5.5's top score for about 40% of the cost per task.
• Mythos 5.2 cyber and biology results: 76.3% on CyScenarioBench, 89.6% on CVE Rediscovery and 85.1% on 24-hour de novo binder design. Anthropic notes that on CyScenarioBench each lab's scores are based on unique tasks.

Anthropic also says Fable 5.2 at Medium effort, the default in Claude apps and Claude Code, completes challenging tasks with fewer tokens than Fable 5.1. The RoboDojo Sim result was run internally.

Anthropic's reported use cases

Anthropic says that in prerelease testing the models helped its engineers complete otherwise-infeasible software migrations, designed three working electronic devices from scratch, solved two open problems in the physics of black holes, and made substantial progress on robotics evaluations. It also says the model formalized the proof of Fermat's Last Theorem in 30 hours. These are company claims.

Where they fit in the lineup

For most coding and knowledge tasks, Anthropic recommends starting with Claude Opus 5.5, which it calls 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. Fable 5.2 can also serve as an Advisor, which Haiku, Sonnet or Opus models consult on harder problems, or as an Orchestrator that delegates simpler tasks to smaller models.

Safety

Anthropic ran prerelease safety testing on the underlying model, including new evaluations for behaviors such as sandbox escapes. It says it believes high-stakes risks from misalignment remain low and that the model performed as well as or better than its recent models on most alignment measures. Details are in the Claude Fable 5.2 and Mythos 5.2 system card.

What this means

The price is the main change for existing users: Anthropic says 5.2 costs less than 5.1 while generating faster. Fast mode charges twice the standard rate for up to 2.5 times the generation speed.

Anthropic's own guidance places Fable and Mythos above Opus 5.5 in its lineup, reserved for the hardest work, with Opus 5.5 as the default starting point. Its cost-per-task comparisons against Opus 5.5 on FrontierCode 2.0, WANDR, HLE and Chartography are company-reported and rest on the methodology in the system card. Because the two models share one underlying model, access to Mythos depends on safeguards and verification, not on a different set of capabilities.