Unverified50% confidenceReleaseExact time
Cognition发布SWE2，基于Kimi K3后训练，官方称在主要评测上与前沿模型得分相当，但成本最多低70%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/11/2026
First Seen
Valid until: 12/10/2026
Sources
Related Entities
Related Claims
UnverifiedKimi 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%63% similarUnverifiedK2.7 Code目前尚未公布SWE-Bench成绩，所有性能数据均为月之暗面自报，此前K2.6曾出现自报成绩与第三方复现相差15-20个百分点的情况63% similarUnverified据官方数据，K3相比K2整体推理效率提升约2.5倍63% similarVerifiedKimi K2.7 Code的思维链token使用量相比K2.6减少约30%62% similarUnverifiedKimi K3前端能力对标GPT 5.6，但API价格只有后者的一半左右61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/908253API
curl https://kongchang.com/api/v1/knowledge/claims/908253MCP
get_claim(id=908253)