Anthropic announces Claude Fable 5.2 and Mythos 5.2 models
Anthropic has introduced new artificial intelligence models, Claude Fable 5.2 and Mythos 5.2, focused on software development and cybersecurity. Along with the performance-enhancing models, API prices have been reduced by 20 percent.
Anthropic has released its new artificial intelligence models, Claude Fable 5.2 and Mythos 5.2. Claude Fable 5.2, which is now available for general use, shows high success in software development processes, while Mythos 5.2 has been restricted specifically to the fields of cybersecurity and scientific research.
The new models promise up to a 30 percent increase in response generation speed compared to the previous generation and up to 2.5 times performance improvement in fast mode. Claude Fable 5.2 managed to outperform the GPT-6 Astra and Gemini 4 Carbon models with a 49.1 percent success rate in FrontierCode 2.0 tests. Mythos 5.2 achieved a score of 76.3 percent in cybersecurity tests and 85.1 percent in protein development tests.
Along with these updates, Anthropic has also changed its API pricing. The fee for 1 million input tokens has been reduced from 10 dollars to 8 dollars, and for output tokens from 50 dollars to 40 dollars, providing a 20 percent discount.
Sources
- cnnturk.com · Link