Unverified50% confidenceBenchmarkExact time
Grok 4系列在多个数学和编程基准测试中跻身第一梯队
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
8/24/2026
First Seen
Valid until: 9/7/2026
Sources
Related Entities
Related Claims
UnverifiedGrok 4.6的智能指数比Grok 4.5高5分，比Grok 4.3高23分63% similarUnverifiedGrok 4.6在UI设计测试中表现平庸，逊于Fable和5.6 Sol63% similarUnverifiedGrok 4.5 在选择最佳已安装 pin 的函数中将语义化版本当作字符串比较，导致 209.9 被误判为大于 209.1163% similarUnverifiedGrok 4.6的综合智能指数评分为61分，基本追平GPT-5.6水平61% similarUnverifiedGrok 4.5 是一个1.5万亿参数的 MoE 模型，此前的 Grok 只有5000亿参数59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/794901API
curl https://kongchang.com/api/v1/knowledge/claims/794901MCP
get_claim(id=794901)