Unverified75% confidenceFactTime unknown

目前Cursor调用的模型上下文长度仅为20万token

1

Sources

75%

Confidence

Medium-term (~90 days)

Relevance

6/1/2026

First Seen

Valid until: 8/30/2026

Sources

AI编程格局深度解析：上下文为何决定Cursor、Windsurf们的生死

bilibiliWhatisNext科技早知道

Related Entities

Related Claims

UnverifiedTiny-MoE模型总参数量约200M，但每个token激活的参数仅约90M69% similar Unverified对于 198B 总参数的稀疏 MoE 模型，实际激活参数量可能仅为 20-40B 量级66% similar Unverified在SWE-bench Full版本上，多数模型的通过率仍低于20%66% similar Unverified34B 80层模型比27B推理时间增加约18%65% similar Unverified1000亿参数的模型处理单次推理请求需约80GB显存，百亿参数级别模型仅需约20GB63% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/25295

API

curl https://kongchang.com/api/v1/knowledge/claims/25295

MCP

get_claim(id=25295)