Unverified80% confidenceFactTime unknown
Kimi K2.5在AMM5 2025综合能力测试中获得96.1分，在GPQA Diamond高难度问题测试中获得87.6分
1
Sources
80%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026
Sources
Kimi K2.5深度测评：性能逼近GPT-5却只要1/4价格
bilibili闲人与杂谈
Related Entities
Related Claims
UnverifiedKimi K2.7与GLM 5.2在FrontierCode Extended基准测试中均取得优异成绩69% similarUnverifiedKimi K2.6在SWE-Bench Pro测试中获得58.6%的分数，位居开源模型第一67% similarUnverifiedKimi K2.7 Code在MCP Mark Verified基准测试上得分81.1，相比K2.6的72.8提升11.4%65% similarVerifiedKimi K2.6采用混合专家（MoE）架构，总参数量达1万亿级别，实际激活参数仅32B64% similarUnverifiedKimi K2.6在SWE Bench Pro、深度搜索问答和Humanity's Last Exam等顶级基准测试上与GPT 5.4、Claude Opus 4.6和Gemini 3.1 Pro等闭源模型竞争64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/24245API
curl https://kongchang.com/api/v1/knowledge/claims/24245MCP
get_claim(id=24245)