Explained improvements to long context quality and memory management in Opus 4.5 and stated that context windows alone are not sufficient.
How media typically covers Dianne Na Penn
Referenced in coverage
Anthropic's Opus 4.5 introduces improved memory management for long-context operations and becomes the first model to score over 80% on SWE-Bench verified coding benchmark.
“Explained improvements to long context quality and memory management in Opus 4.5 and stated that context windows alone are not sufficient.”