Expired50% confidenceBenchmarkExact time
GPT-5.6's Extreme High tier achieves a score of 53.6% for approximately $63, whereas Claude Sonnet in adaptive mode spends approximately $2,000 to achieve a score of only 40.5%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/13/2026
First Seen
Valid until: 10/11/2026(expired)
Sources
GPT-5.6发布：三款模型对标Claude，ChatGPT Work超级应用全解析
bilibiliAI-seeker7/9/2026
Related Claims
UnverifiedGPT-5.6 set a new all-time high score of 53.6 on the Agent's Last Exam benchmark, surpassing Claude Opus 4.5 by 13.1 points.71% similarUnverifiedIn SWE-bench testing, the early GPT-4 resolution rate was below 2%, while Claude 3.5 Sonnet broke through 49%.70% similarVerifiedClaude Sonnet 5 scores 63.22% on SWE-bench Verified (Agentic Coding), approximately 6 percentage points behind Opus 4.8.69% similarUnverifiedIn testing, GLM 5.1 achieved a composite score of 89.3, ranking first and surpassing Claude Sonnet 4.6's score of 87.2.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/501553API
curl https://kongchang.com/api/v1/knowledge/claims/501553MCP
get_claim(id=501553)