deepseek-v4.1IMAGE
LobsterAI report: V4.1 Flash cached input costs 4.6% of list price
LobsterAI's September report says 73.6% of ordinary users rely on a single model and 77% issue AI tasks via WeChat, noting DeepSeek V4.1 Flash cached input costs just 4.6% of the original price.
Modengsir AI
@ModengSirArchived 12 days agoOriginally posted 2026.10.19
Modengsir AI @ModengSirOriginally posted 2026.10.19
AI 圈内沉迷跑分、Token 定价， 很容易困在信息茧房。 LobsterAI 9 月报告几组反常识数据： ✅73.6% 普通用户只用一个模型 ✅77% 在微信下达 AI 任务，远超飞书 ✅缓存大幅降本：DeepSeek V4.1 Flash 缓存输入成本仅原价 4.6% 技术的终点是傻瓜模式：少折腾，拿结果。 你身边不追 AI 热点的朋友，平时怎么用 AI？
This post quotes another post: “Office agent report: V4.1 Flash cached input costs just 4.6% of list price”
x.com/modengsir/status/2112022079273087241Archived 2026.10.19 12:05