Anthropics new Claude 4.1 dominates coding tests days before GPT-5 arrives

Anthropic's latest model, Claude Opus 4.1, has set a new benchmark by achieving a 74.5% score on coding evaluation tests, positioning it as a leader in AI coding capabilities. However, despite this technical advancement, the company's revenue model faces significant risk, as nearly 50% of its $3.1 billion API revenue is concentrated among just two major customers, highlighting potential vulnerabilities in its market diversification.