Subquadratic launches with a $29M seed and debuts SubQ, an LLM that uses a subquadratic sparse attention architecture to achieve a 12M-token context window (Kyt Dotson/SiliconANGLE)

Subquadratic has launched with $29M in seed funding and introduced SubQ, an LLM featuring a subquadratic sparse attention architecture that achieves a 12M-token context window—dramatically expanding the amount of text an AI model can process simultaneously, which could reduce costs and improve performance for enterprise applications requiring extensive context understanding. This advancement addresses a critical bottleneck in LLM deployment, enabling IT organizations to handle significantly larger documents, code repositories, and conversational histories without the computational overhead of traditional transformer models. For technology leaders, this represents an emerging alternative to current foundational models that could reshape decisions around AI infrastructure, vendor selection, and the feasibility of deploying sophisticated AI solutions across resource-constrained environments.