SubQ Snap 决策模型实现线性扩展，384K 上下文 660ms 起步
Subquadratic 出了个决策模型 SubQ Snap，384K token 输入只要 2.5 秒，延迟基本线性增长，输入每百万 token 5 美分，输出还免费。
Subquadratic 推出的决策模型 SubQ Snap 在长上下文上接近线性扩展：100K token 延迟 660ms，384K token 延迟 2.5 秒，输入量增约 3.84 倍耗时约 3.8 倍。若用标准二次方注意力，同区间注意力开销会上升约 15 倍。这让单次调用可容纳 384K token 的代码、文档、日志和截图，而多数模型上限在 32K 到 64K，需要分块处理。定价为输入每百万 token 0.05 美元，输出免费。此前的决策模型多为 32K 上下文的短文本工具，难以对照整个代码库做判断。
Decision models deserve more attention than they get.
look at the latency numbers on SubQ Snap from @subquadratic
660ms for 100K tokens. 2.5 seconds for 384K. that’s about 3.8x the time for 3.84x the input, which is what linear scaling looks like. with standard quadratic attention, the attention cost alone would rise about 15x over that range.
that scaling is how a decision model can hold 384K tokens of code, docs, logs and screenshots in one call, when most stop at 32K to 64K and make you chunk.
$0.05 per million input tokens, output free.
many of the existing decision models have mostly been short-context tools. 32K tokens is enough to classify a ticket, but too little to judge a change against a whole codebase. Now that changed.
