Claude Fable 5.2 and Mythos 5.2 cement Anthropic's AI lead just before its IPO
Ahead of its initial public offering, Anthropic has launched Claude Fable 5.2 and Claude Mythos 5.2, aimed at demanding technical workloads while keeping Opus 5.5 as its recommendation for everyday use. Both models share a common foundation but diverge in access: Fable 5.2 is open to the public with built-in safeguards that block biology and cybersecurity tasks, while Mythos 5.2 is reserved for vetted organizations in cybersecurity and life sciences. Base pricing dropped to $8 per million input tokens and $40 per million output tokens, delivering output up to 30 percent faster than their version 5.1 predecessors.
On Artificial Analysis’s Index v5 benchmark, Fable 5.2 captured first place at the highest compute tier with 65 points, ahead of Gemini 4 Carbon (60) and Opus 5.5 (59). It achieved top marks across specialized tests, scoring 80 percent on HLE-Diamond and 73 percent on Terminal-Bench-Science, though highest-tier tasks average $26.50 each. Meanwhile, Mythos 5.2 demonstrated notable security gains, discovering 10.2 percent more vulnerabilities than Opus 5.5 in an October 2026 test across 1,829 CVEs, with Anthropic noting it deliberately constrained features capable of accelerating recursive self-improvement.
Continue in the app — vote & join in ➔