The author tested both models via API and found that Claude Fable 5.2 and Opus 5.5 scored the same across nine coding, agentic, and troubleshooting tasks. However, their estimated total costs were $8.98 and $2.29, respectively, with total runtimes of 1,579 seconds and 701 seconds.
Source:The New Stack · Read original
Leave a Reply