Anthropic ships Mythos and Fable models with no benchmarks given

single source· 1 articles · confidence: low · first seen 2026-11-09 19:24 UTC

What this means for you

Nothing to act on yet: no pricing, no availability date and no published evaluation. If you are picking a model this month, this release gives you no basis to switch — wait for scores that arrive with an evaluation date and a described test setup.

Anthropic released two models, Mythos and Fable, on 9 November, saying they improve on coding, reasoning and vision and communicate more clearly with users. The coverage carries no benchmark scores, no evaluation dates, no prices and no availability details, so the claims cannot yet be checked. TechCrunch frames the release around concerns about recursive self-improvement — systems contributing to the design of their own successors. Nothing in the report says how the two models differ from each other, or from Anthropic's existing ones.

Key facts

  • ·Anthropic released two models named Mythos and Fable on 9 November 2026. source
  • ·Anthropic says the models show improvements in coding, reasoning and vision. source
  • ·Anthropic says the models communicate more clearly with users. source
  • ·The report gives no benchmark scores, evaluation dates, prices or availability details. source
  • ·TechCrunch frames the release around concerns about recursive self-improvement (RSI). source

What the sources say

  • TechCrunch AI — Carries the company's claims only; no scores, prices or availability dates appear.

Sources

The original reporting. Follow these — they did the work.

← the wire