· ai-devtools
Anthropic launches Fable 5.2, Mythos 5.2 with price cut, faster speed, better coding scores
Anthropic has released Claude Fable 5.2 and Mythos 5.2, the newest iterations of its flagship models.
Fable 5.2 is immediately accessible across platforms like Amazon Web Services, Google Cloud, and Microsoft Azure.
While both models share a core architecture, Fable 5.2 incorporates stricter safeguards concerning biology and cybersecurity. The company recommends Opus 5.5 for most knowledge and coding tasks, positioning Fable 5.2 to help users "tackle the most daunting technical challenges."
A notable enhancement is speed; Anthropic states that Fable 5.2 and Mythos 5.2 generate output up to 30 percent faster than their 5.1 predecessors. Furthermore, a Fast mode is now available for both, which accelerates token generation by up to 2.5 times.
Regarding cost, Anthropic is reducing the price for Fable, lowering it from $10/$50 per million input/output tokens to $8/$40. Mythos 5.2 shares these same rates. Even with this reduction, Fable's standard rates are twice those of Opus 5.5, though Anthropic suggests this doesn't equate to a higher cost per task.
In coding benchmarks, Fable 5.2 showed significant gains, scoring 49.1 percent on FrontierCode 2.0, up from 32.4 percent for Fable 5.1. On FrontierSWE v2, it achieved 69.9 percent, outperforming Astra's 65.5 percent and Opus's 62.3 percent.
The models also exhibit improved speed and fewer safety interruptions. Early testing showed roughly 55 percent fewer overall blocks across Claude.ai, Claude Code, and Cowork, with cybersecurity blocks dropping by about 60 percent.
In scientific applications, Fable 5.2 reportedly solved two open theoretical-physics problems related to black holes and holographic duality. Additionally, a team of 1,000 Fable 5.2 agents formalized the proof of Fermat's Last Theorem in Lean in 30 hours.
For API developers, Anthropic is extending preserved thinking to all accounts, a feature that restricts modifications to Claude’s prior context to protect its reasoning for distillation.
Why it matters
For developers, this release provides more performant and cost-effective options for tackling complex technical challenges. The improved coding scores and speed enhancements directly impact productivity when using these models for development tasks.
