Unverified50% confidenceBenchmarkExact time
开发者宣称MiAPI在SimpleQA基准测试上达到82%的准确率并超越Perplexity
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/12/2026
First Seen
Valid until: 12/11/2026
Sources
Related Entities
Related Claims
UnverifiedEmsyAI使用FP16混合精度训练了15000步，验证困惑度收敛到5.8668% similarUnverified4Bit量化下模型的基准测试得分通常能保持原始精度的90%-95%67% similarUnverifiedQuerit团队使用FreshQA数据集进行性能评估，准确率达到83.17%67% similarUnverifiedNVFP4量化方案相比BF16基线实现了1.5倍的推理加速，同时在top-1准确率上保留了92%到97%的水平66% similarUnverifiedPrecision Mode 在一组涵盖约9,000份复杂文档的基准测试中达到了94.7%的提取准确率66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/914759API
curl https://kongchang.com/api/v1/knowledge/claims/914759MCP
get_claim(id=914759)