卖什么
LLM context compression for better accuracy
商业模式拆解
买家：谁掌握预算？
未核实AI/ML engineering lead or head of platform at companies with large LLM token spend.
触发点：什么让他们现在购买？
High LLM bills and latency from long prompts that carry far more tokens than the model uses.
工作流：什么输入变成什么输出？
pip install the SDK, send long context plus the query to the compression API, get back a query-aware compressed context, then pass it to the LLM; includes cost receipt and playground.
经济账：省了什么，怎么收费？
Value claim: up to 90% fewer tokens, avg 10x compression, up to 226x in demo, +3.7pp accuracy uplift; example Boeing 10-K query $0.263 to $0.037.
渠道：怎么接触到客户？
Developer-led: pip SDK, docs and playground, with book-a-call for enterprise; cloud API or on-prem in customer VPC.
融资历史与报道
| 融资日期 | 轮次 | 金额 | 投资方 | 来源 |
|---|---|---|---|---|
| — | YC 2026 冬季 | — | Y Combinator | ycombinator.com |
档案汇总自公开报道和公司官网。商业模式部分是 MaxBC 对这些材料的解读，不是已确认的事实；未知之处留空，不作猜测。 更新于 2026-10-12.