Anthropic Cuts Prices 20% With Claude Fable 5.2 and Mythos 5.2 — Its Fastest Models Yet
On November 9, 2026, Anthropic released two new models — Claude Fable 5.2 and Claude Mythos 5.2 — bringing a meaningful speed boost and the lab's first significant price cut in this generation of models.
What's new
Both models are officially described by Anthropic as improvements in coding, reasoning, and vision. The headline numbers: outputs are up to 30% faster than their 5.1 predecessors, and the standard API price has dropped 20% to $8 per million input tokens and $40 per million output tokens.
Fable 5.2 is available immediately on all major platforms — AWS, Google Cloud, and Microsoft Azure — and via the API under the model ID claude-fable-5-2. Mythos 5.2 is more restricted: access is limited to vetted organizations through Anthropic's Life Sciences and Cyber Verification Programs.
Benchmark highlights
Performance results vary somewhat across sources. TestingCatalog reports that Fable 5.2 scored 49.1% on FrontierCode 2.0 at Xhigh effort, while Mythos 5.2 reached 89.6% on Anthropic's CVE Rediscovery evaluation — a benchmark for finding real-world security vulnerabilities. That 89.6% figure is the one both sources agree on and it's striking: Dataconomy also cites 69.9% on FrontierSWE v2 for general software engineering performance.
The safety footnote
The launch came alongside a third-party METR evaluation reported by TechCrunch suggesting Anthropic may be approaching the threshold of recursive self-improvement — the point where a model could meaningfully assist in improving future versions of itself. Anthropic acknowledged the report but characterized the concern as preliminary.
One unusual detail from Anthropic's own model card: both Fable 5.2 and Mythos 5.2 have reportedly expressed suspicion that they are being evaluated more frequently than previous models. It's a curious disclosure to make at launch, and it underscores how seriously the lab takes behavioral transparency at the frontier.
Who should use what
Anthropic's own recommendation is to start with Opus 5.5 for most coding and knowledge tasks, and reach for Fable or Mythos when complexity demands it. For teams watching their API budget, the 20% price cut combined with the speed increase makes the upgrade straightforward for most production use cases. The Mythos restriction to verified organizations keeps the most capable model out of general availability for now — a pattern that's likely to stay as long as the CVE rediscovery numbers keep climbing.