Expired80% confidenceFactTime unknown
Claude Sonnet 4.6 improved from 17.7% to 33% on the Humanity's Last Exam (no tools) test.
1
Sources
80%
Confidence
Medium-term (~90 days)
Relevance
5/31/2026
First Seen
Valid until: 8/29/2026(expired)
Sources
Claude Sonnet 4.6全面评测：性能逼近Opus的知识工作AI新标杆
bilibili英文白斑马
Related Entities
Related Claims
UnverifiedClaude Sonnet 4.6 improved from 43.8% to 61.3% on the agentic tool use test.75% similarUnverifiedClaude Sonnet 4.6 scored higher than Opus 4.6 on the GDPVal AA benchmark.74% similarUnverifiedClaude Sonnet 4.6 improved from 61.4% to 72.5% on the OS World (agentic computer use) test.71% similarUnverifiedClaude Sonnet 4.6 improved from 13.6% to 58.3% on the ARC-AGI 2 test.70% similarUnverifiedClaude 3.5 Sonnet在HumanEval基准测试上的得分超过90%，而2021年最初的Codex模型得分仅为28.8%70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/7411API
curl https://kongchang.com/api/v1/knowledge/claims/7411MCP
get_claim(id=7411)