Verified80% confidenceFactTime unknown
On mainstream evaluations such as MMLU, HumanEval, and GSM8K, the score gaps among leading large models are often only a few percentage points.
4
Sources
80%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
DeepSeek V4 Pro优惠期延长至2026年5月：开发者成本利好全解读
twitterdeepseek_ai
Related Entities
Related Claims
UnverifiedMultiple academic studies have shown that models scoring highest on mainstream benchmarks such as MMLU and HumanEval often perform significantly worse on novel questions of equivalent difficulty.77% similarUnverifiedMid-tier models typically achieve more than 80% of flagship model scores on benchmarks such as MMLU and HumanEval, while requiring less than 30% of the inference cost.75% similarUnverifiedThe score gap among multiple frontier models on MMLU has narrowed to the order of one percentage point, driving the emergence of more challenging benchmarks such as MMLU-Pro and ARC-AGI.72% similarUnverified斯坦福大学HAI研究所2024年度AI指数报告显示，顶级闭源模型在MMLU基准上的最高分与第五名之间的差距，已从2022年的约8个百分点收窄至不足2个百分点72% similarUnverifiedThe tester concluded that only a difference of approximately 5% or more would be statistically meaningful on a dataset of this size, and since the actual gap is under 4%, it has not reached a significant level.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/27923API
curl https://kongchang.com/api/v1/knowledge/claims/27923MCP
get_claim(id=27923)