Anthropic's most capable model, codenamed "Model 2," is for internal use only
decoder ·
Anthropic is reportedly using an unreleased AI model internally, codenamed "Model 2," which outperforms all publicly available Claude versions, according to the company's August 2026 Risk Report. The model is classified in the Mythos class and is described as slightly stronger overall than Claude Mythos 5, though it is weaker in some areas and does not represent a major capability leap. On Anthropic's internal capability index (AECI), Model 2 scores about 1.5 points above Mythos 5, a smaller gain than the jump from Mythos Preview to Mythos 5. Internally, Anthropic relies heavily on the model for coding, data generation, and research, with Claude now writing most of the code in production systems. Model 2 underwent an internal review before deployment but was not tested as thoroughly as Mythos 5, and no new or more serious misalignments were found, with overall misalignment risk rated as low. The company currently has no plans to release Model 2 externally.
decoder