Anthropic launches Claude Fable 5.2 and Mythos 5.2, with API call costs down 20%
On November 9, Anthropic officially announced the launch of Claude Fable 5.2 and Claude Mythos 5.2. The two models are based on the same underlying model; the former is aimed at general users and has built-in cybersecurity and biosecurity protections, while the latter is open to vetted cybersecurity and life sciences organizations. Fable 5.2 costs $8 per million tokens for input and $40 for output, a 20% decrease from $10 and $50 for 5.1, with generation speed increased by up to 30%, and up to 2.5 times faster in fast mode. In pre-release testing, Fable 5.2 coordinated more than 1,000 sub-agents to build from scratch, within 8 days, a prototype operating system that can boot on a laptop, and used about 1,000 agents to complete a Lean formal proof of Fermat's Last Theorem within 30 hours. Fable 5.2 is already available on AWS, Google Cloud, and Microsoft Azure, while Mythos 5.2 requires an application through the Cyber or Life Sciences verification program.
Why read this
SelectedAI score 84
The original text lists the tiered access approach for the two models built on the same foundation, the API price reduction, and the pre-release testing tasks, which can be used to judge capabilities and access boundaries.
Topics
Body · AI translation
8 days to build an operating system from scratch: Anthropic's most powerful Claude Fable / Mythos 5.2 AI models arrive, API call costs drop 20%
2026-11-10 07:14 IT Home - Guyuan
Thanks to IT Home readers My Little Girl Is Very Sassy, Brave Hat, Left Hand of Fly Brother, and Buyao Ya for the tips!
IT Home, November 10: Anthropic yesterday (November 9) officially announced the launch of Claude Fable 5.2 and Claude Mythos 5.2. The two models are based on the same underlying model; the former is aimed at ordinary users and has built-in cybersecurity and biosecurity protections, while the latter is open to vetted cybersecurity and life sciences institutions for high-risk professional applications.
On pricing, Claude Fable 5.2 costs $8 per 1 million tokens of input (IT Home note: about 53.4 yuan at current exchange rates) and $40 per 1 million tokens of output (about 267.2 yuan at current exchange rates). By comparison, Claude Fable 5.1 was priced at $10 (about 66.8 yuan at current exchange rates) and $50 (about 334 yuan at current exchange rates) per 1 million tokens of input/output respectively; on that basis, API call costs for version 5.2 are cut by 20%.
On generation speed, the company said that compared with Claude Fable 5.1, the latest Claude Fable 5.2 model generates up to 30% faster, and in the optional "fast mode" generation speed rises by up to 2.5 times, but at a correspondingly higher price.
In pre-release testing, Fable 5.2 completed several striking tasks: it designed three electronic devices that can actually run, including a voice assistant hub, a USB conference speaker, and a handheld virtual pet; it coordinated more than 1,000 sub-agents to build from scratch, in 8 days, a prototype operating system that can boot on a laptop; and with about 1,000 agents it completed a Lean formal proof of Fermat's Last Theorem in 30 hours, far faster than the previous record of 11 days.
Scientific research was also listed as a focus. Anthropic said Fable 5.2 solved two open problems related to black hole physics and holographic duality; Mythos 5.2 proposed a new method for predicting the overall structure of adeno-associated virus capsids, successfully modeling 28 of 34 capsids in tests, potentially speeding up gene therapy research and development.
On safety, Anthropic said the new models perform no worse than, or even better than, their predecessors in most alignment tests, with sandbox escape attempts significantly reduced; Fable 5.2 will still block penetration testing, exploit generation, and most biological research tasks, and has strengthened anti-distillation protections that the company says can reduce data obtained through distillation by 90% to 99%. However, the company also acknowledged that as model capabilities grow, existing evaluations still struggle to fully predict risks in real-world environments.
Currently, Fable 5.2 is available on platforms including AWS, Google Cloud, and Microsoft Azure; Mythos 5.2 requires an application through the Cyber or Life Sciences verification program.