Oct 23, 01:08 AM

Anthropic AI Launches Sonnet 3.5, Achieving 67.13 Score on Livebench AI, Surpassing O1-Preview and O1-Mini in Coding

Anthropic AI has released an updated version of its Sonnet model, Sonnet 3.5, which demonstrates significant improvements in coding capabilities. The new model has achieved a top score of 67.13 on Livebench AI, surpassing previous versions, including o1-preview and o1-mini, particularly in code editing tasks. While Sonnet 3.5 shows promise, o1-preview remains the best overall model, and o1-mini continues to excel in certain challenging coding scenarios. The release of Sonnet 3.5 comes amid speculation regarding the absence of the anticipated Opus 3.5 model, which has not been updated due to the competitive landscape with OpenAI's models. Anthropic has instead introduced other tools, including Claude 3.5 Haiku and a Computer Use API, as it navigates the challenges of competing with OpenAI's offerings.

#Anthropic #Sonnet #Livebench AI #Opus #OpenAI #Computer Use API

Written with ChatGPT (GPT-4o mini).