Unverified50% confidenceBenchmarkExact time
In internal hard-problem tests, Kimi K3.1 Pro's hallucination rate dropped from K3's 23.4% to 8.3%.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/24/2026
First Seen
Valid until: 11/7/2026
Sources
Related Entities
Related Claims
UnverifiedKimi K3 performed very poorly in this test, and the UP host suspects it has recently undergone a quality degradation.78% similarUnverified以Kimi K3实验，给它预填充1%的Claude Opus解码推理片段，其输出风格就会明显向Opus靠拢，某些区段Claude推理链从Kimi K3提取的容易程度比其他模型高出约6个数量级69% similarUnverifiedKimi K3.1 Pro scored 92.7 on 3D modeling and 79.6 on computer operation.69% similarUnverifiedIn LSCBench, Kimi-K3 had the highest functional pass rate (91.3%), but the lowest functional + security pass rate (16.4%).69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1092242API
curl https://kongchang.com/api/v1/knowledge/claims/1092242MCP
get_claim(id=1092242)