卖什么
AI models that learn from experience
商业模式拆解
触发点：什么让他们现在购买？
Science and engineering progress gated by the cost of physical experiments (drug and materials discovery).
工作流：什么输入变成什么输出？
Research approach: LLMs write code (program synthesis) to model a task from few examples; reached 97.9% on ARC-AGI-2 at about $12 per task; aims at hypothesis generation and data-efficient modeling for hardware engineering, drug and materials discovery.
经济账：省了什么，怎么收费？
未核实No product pricing; pre-commercial research lab, partnerships with researchers and hardware/biology/materials teams.
渠道：怎么接触到客户？
Seeking research and engineering partnerships/collaborators via founders email; open-sourced solver.
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.