SubQ Releases Its 1.1 Small Model Card as Dangel and Whedon Try to Prove Long Context Can Beat RAG

The Subquadratic team says its sparse-attention model hits 98% needle retrieval at 12M tokens, but access remains limited to design partners.

By · Published

Primary source: Subquadratic

Why it matters

SubQ is attacking one of the most expensive assumptions in enterprise AI: that large artifacts must be chunked, retrieved and reconstructed because models cannot afford to reason over them directly.

Illustration of SubQ's 1.1 Small model card floating over a field of data tokens with a single glowing needle highlighting a retrieved item.

Subquadratic co-founders Justin Dangel and Alexander Whedon are moving SubQ from a provocative May launch claim into a more testable phase: the company on June 16th released the model card and technical report for SubQ 1.1 Small, a long-context model built around what it calls Subquadratic Sparse Attention.…

Reader comments

Conversation for this story loads after sign-in.