Verified70% confidenceBenchmarkExact time
On mainstream benchmarks such as MMLU and HumanEval, the score gaps among Claude 3.5, GPT-4o, Gemini 1.5 Pro, and Grok-2 have been compressed to within the range of statistical error.
4
Sources
70%
Confidence
Medium-term (~90 days)
Relevance
7/24/2026
First Seen
Valid until: 10/22/2026
Sources
Anthropic放弃订阅制转向按量计费，是明智还是自断后路？
redditr/ClaudeAI7/12/2026
Related Claims
UnverifiedIn EvoArena testing, GPT-5.5, which achieved the highest Step Level score, had an accuracy of just over 50%, while Gemini-3.1 Pro and GLM-5.1, which performed best at the Chain Level, reached only approximately 28%–29%.71% similarUnverifiedModels like GPT-4 and Claude 3.5 surpassed 90% accuracy on MMLU, significantly diminishing its discriminative power69% similarUnverified2024-2025年间，Claude 3.5 Sonnet、GPT-4o、Gemini 2.0等模型在HumanEval、SWE-bench等编程基准测试的通过率从不足50%提升至90%以上69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/608322API
curl https://kongchang.com/api/v1/knowledge/claims/608322MCP
get_claim(id=608322)