Unverified50% confidenceBenchmarkExact time
All evaluated model-harness configurations achieved pass rates of no more than 30% on Harbor-Index
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/11/2026
First Seen
Valid until: 12/10/2026
Sources
Related Entities
Related Claims
UnverifiedEven the best-performing models can only recover 30% of the evaluated evidence and 25% of the limitations descriptions.63% similarVerifiedFor models with more than 30B parameters, the performance loss caused by 4-bit quantization typically results in a 1–3% drop on benchmarks, but the impact is more significant for smaller models.60% 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.59% similarUnverified在SWE-bench Full版本上，多数模型的通过率仍低于20%59% similarUnverifiedThe proportion of enterprises that can clearly quantify the ROI of AI investment still remains below 30%.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/902673API
curl https://kongchang.com/api/v1/knowledge/claims/902673MCP
get_claim(id=902673)