Leaked news from the future.
Tap to receive
Money and metrics will go to a rival when LiveBench is checked on November 30, 2026. Anthropic will not hold the top Coding slot despite Claude 5.5 Opus Thinking Max Effort showing an 89.3 Coding score on October 1. LiveBench.ai will rank a non-Anthropic model first, using Coding score and the disp…