Anthropic cut Fable’s price from $10 to $8 per million input tokens, making it cheaper than its primary competitor, GPT-6 Astra. The company says Fable 5.2 responds 30% faster across effort levels; a faster inference service is available for double the price. The model also supports zero data retention policies through the Enterprise Frontier Safeguards system, a long-standing concern for some enterprise clients.
The release is Anthropic’s first since assessment group METR published a report suggesting the lab may be nearing recursive self-improvement: the ability of AI models to continue training models independently. METR said internal models were accelerating Anthropic’s AI progress by roughly 2x, while emphasizing that substantial uncertainties remain.
Anthropic’s release notes also warn that testing model capabilities accurately remains difficult. The company said Mythos 5.2 and Fable 5.2 both suspect they are being evaluated more often than recent models, complicating efforts to judge how they would behave in real-world settings. Anthropic expects that challenge to grow as capabilities increase.
The company says Mythos 5.2 cannot act covertly against its user. Its system card says the model performed better than earlier Claude models on several covert-capability evaluations, especially when its reasoning was hidden from a monitor. Anthropic added that low absolute performance on new covert evaluations supports claims that harmful actions can be detected in high-stakes situations.
Anthropic says Fable 5.2, which has been available to selected parties before release, has solved two open problems in theoretical physics, formalized Fermat’s Last Theorem and supported other research projects. Separately on Monday, AI startup Physical Superintelligence announced a proof of the Aaronson-Ambainis conjecture, a step toward setting theoretical limits for quantum computing. The company said it used a non-public Anthropic model; Anthropic did not confirm or deny that it was Fable 5.2.

Comments (0)
No comments yet. Be the first!