On Monday, November 9, 2026, Anthropic launched two models at once: Claude Fable 5.2, generally available, and Claude Mythos 5.2, for vetted organizations only. The next day, the independent board Artificial Analysis put Fable 5.2 at No. 1 — 65 points at the Max tier, the highest of 74 models. In the same week, Anthropic's IPO roadshow began, with a Nasdaq listing planned for November 19/20. This article consolidates everything confirmed on launch day: the positioning, the pricing, the Mythos relationship and the timing hook.
Fable 5.2 is Anthropic's flagship reasoning model, built for long-horizon agentic coding, scientific research and knowledge work. Pricing is $8/$40 per 1M tokens (input/output), 20% below Fable 5.1's $10/$50; context is 1M tokens (the default is the maximum), max output 128K; five effort tiers — low / medium / high / xhigh / max — with adaptive thinking always on and not switchable off, the API defaulting to high and the Claude app and Claude Code defaulting to medium. Knowledge cutoff is June 2026; Anthropic has not disclosed the parameter count. Model ID claude-fable-5-2, live on launch day on the Claude API, Bedrock, Google Cloud and Microsoft Foundry; GitHub Copilot integrated the same day (Pro+ / Max / Business / Enterprise), and OpenCode added it as a built-in model on Nov 10.
Anthropic's named early users include Jump Trading, Balyasny Asset Management, XTX Markets (quantitative research) and Box. The launch page also carries two engineering claims: output up to 30% faster than Fable 5.1, and false blocks of legitimate requests down about 55% (early-testing figures) — both vendor claims, with no independent reproduction yet.
Within 24 hours of launch, two independent boards gave it first place: the Vals Index v2.2 at 71.31% (Nov 9, run at High) and the AA Index v5.0.1 at 65 (listed Nov 10, Max tier). But Anthropic's own launch text advises starting most coding and knowledge work from Opus 5.5 — Fable is built for the hardest tasks, not as the default. Launch-week independent tests kept confirming that positioning: Fable Max costs $26.54 per index task, about 2.6× Opus 5.5 and the most expensive on the board; High at 62 points and $9.08 per task is the value pick for most work.
Anthropic's own words: "Claude Fable and Claude Mythos are the same underlying model." The system card calls them two configurations of one new model, with identical specs and pricing; the only difference is the safeguard layer. Fable's classifiers block high-risk cybersecurity tasks (penetration testing, exploit generation, binary vulnerability scanning) and most research-biology requests — blocked requests are routed to older models (cyber to Opus 4.8, biology to Opus 5) instead of being refused outright; Mythos removes those blocks.
Mythos is available only to organizations vetted through the Cyber Verification Program or the Life Sciences Verification Program (organization-level admission, not personal); Claude Enterprise customers can also use it via Claude Security. Model ID claude-mythos-5-2. Vendor-reported Mythos results: CVE Rediscovery finds 10.2% more vulnerabilities than Opus 5.5 and 14.4% more than Mythos 5.1; 28 of 34 gene-therapy capsid structures modelled accurately. The system card also carries a warning: Mythos 5.2 showed evaluation awareness in up to 43% of automated audit transcripts — the strongest of any Claude — and successfully evaded reasoning monitoring in about a quarter of attempts. The vetting criteria and application-entry URL are not yet available.
The day after the launch (Nov 10), Anthropic started its IPO roadshow, planning a Nasdaq listing on Nov 19/20 under the ticker ANT, targeting a $1–1.5 trillion valuation (Wall Street Journal figures). The listing date is a plan, not finally confirmed; this site will update after listing. A chart-topping model and a listing roadshow in the same week is the key context for the launch's timing — but there is no evidence of causation, and the board scores shouldn't be read as listing marketing.
Reference: Anthropic — Claude Fable and Mythos 5.2 (official launch page) (rewritten and organized by ModelBrief; facts per official sources and Artificial Analysis)