claude-opus-4-20250514· Anthropic LMArena Human Preference 140Kcaculate 9.9-9.11 Save Share Download ReportFrom a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with qwen3-235b-a22b-no-thinking’s answer. UserCopy promptcaculate 9.9-9.11 AssistantCopy response9.9 - 9.11 = -0.21 The result is negative 0.21.Copy first promptCopy first responseOther models, same or similar promptResponses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.Identical first promptqwen3-235b-a22b-no-thinkingCompare side by side →