Claude 4 Tested: Why It Might Outperform ChatGPT in the AI Race
Claude 4 from Anthropic sets new records: 72.5% on SWE-bench and 7 hours of autonomous work. We may have the best AI model for coding. And that’s not all.Alex Albert from Anthropic didn’t hold back: “Claude 4 is the best programming model in the world.” A bold claim — but the data seems to...