Anthropic Launches Claude Fable 5.2 and Mythos 5.2, Cuts Prices 20%
3 min readAnthropic released Claude Fable 5.2 and Claude Mythos 5.2 on Monday, pairing its best coding and reasoning scores to date with a 20 percent price cut. The new models arrive two months after Fable 5.1 and as the company prepares its public listing, which makes this launch as much a business statement as a technical one.
Background
Fable and Mythos share one underlying model. Fable is the generally available version and carries safeguards that route risky cyber and biology requests to less capable models. Mythos is the unrestricted version, limited to organizations vetted through Anthropic’s Cyber and Life Sciences verification programs. Fable 5.1 shipped in September at $10 per million input tokens and $50 per million output tokens, and OpenAI’s GPT-6 Astra sells at the same rates, according to Fortune.
What Happened
According to Anthropic, Fable 5.2 scores 49.1 percent on FrontierCode 2.0, up from 32.4 percent for Fable 5.1 and ahead of Opus 5.5 at 45.5 percent and GPT-6 Astra at 42.8 percent. It posts 74.2 percent on Humanity’s Last Exam with tools and 73.4 percent on the Chartography vision test, a gain of almost 30 points over 5.1. Mythos 5.2 reaches 89.6 percent on CVE Rediscovery, Anthropic’s vulnerability-finding evaluation.
Pricing drops to $8 per million input tokens and $40 per million output tokens, with cache reads at $0.20. Fortune notes that puts Anthropic 20 percent under GPT-6 Astra on standard tokens and as much as 80 percent under it on cached input. A new fast mode generates tokens up to 2.5 times faster at double the price. MacRumors reports output is up to 30 percent faster than 5.1 and that the model triggers about 55 percent fewer safeguard blocks. Fable 5.2 is live on the Claude Platform, AWS, Google Cloud and Microsoft Foundry, and for Pro, Max, Ultra, Team and Enterprise subscribers.
Anthropic also showcased long-horizon agent work. Roughly 1,000 Fable 5.2 agents produced a computer-verified formalization of Fermat’s Last Theorem in 30 hours, and Uber ran a security audit with about 170 reviewer agents.
Why It Matters
The price cut lands as OpenAI’s Codex has been winning developers on cost and Anthropic’s latest S-1 showed third-quarter growth slowing. Cheaper tokens plus a coding lead is a direct pitch to the enterprise accounts that drive its revenue. The split between Fable and Mythos also shows how Anthropic plans to ship frontier capability while keeping the most dangerous uses behind a verification gate. Watch whether OpenAI matches the price, and whether Mythos 5.2 access widens beyond security and life sciences teams.
For developers the upgrade is simple: the model ID is claude-fable-5-2, and this time the price went down rather than up.