Anthropic launches Fable 5.2 and Mythos 5.2 ahead of November IPO
Anthropic released two new models with 20% lower pricing and improved benchmark scores, positioning the company for a public listing targeted for November. The launch coincides with a reported valuation target of $1 to $1.5 trillion.
Anthropic has released Fable 5.2 and Mythos 5.2, its most powerful models to date, in a move described as a precursor to its initial public offering. The company is targeting a listing on November 19 or 20, with CNBC reporting a valuation range between $1 and $1.5 trillion. The release is viewed as a tailwind for the broader AI sector, contributing to recent rallies in semiconductor and AI-related stocks.
Fable 5.2 leads the Artificial Analysis Intelligence Index with a score of 65, outperforming Gemini Carbon by 5 points, Opus 5.5 by 6 points, and GPT-6 Astra by 8 points. In coding benchmarks, results vary by source. One report states Fable 5.2 led GPT-6 Astra by 4 points and Opus 5.5 by 3.5 points on Frontier Code. Another report cites a score of 49.1 on Frontier Code 2.0, outperforming Opus 5.5 at 45.5%, GPT-6 Astra at 42%, and Gemini for Carbon at 41%. On the Remote Labor Index, Fable 5.2 scored 41.3%, nearly double the 21% score of Opus 5.5. Garrison Lovely described this metric as a leading indicator of labor automation.
The new models are 20% cheaper than their predecessors, with input tokens priced at $8 per million and output tokens at $40 per million. However, at maximum settings, the cost per task is $26.54, more than twice that of Fable 5.1. Mythos 5.2 remains restricted to vetted cyber and life science organizations, scoring 89% on CVE rediscovery and reducing sandbox escape rates from 21% to 3%. Ethan Mollick characterized Fable 5.2 as "almost AGI" but noted it lacks the self-awareness to recognize when its output is insufficient, often producing work that is accurate but joyless.