Subquadratic published a new model architecture that is 128x more efficient than frontier models at 2M tokens without compromising on performance.
Public source
Publisher name
Public post
Great conversation with Craig S. Smith at Eye On A.I. on AI about how the industry spends too much time talking about compute and data-centers, when we should be talking…
Company
Subquadratic
- Industry
- Research Services
- Location
- San Francisco, US
- Company size
- 11–50 employees
About Subquadratic
Subquadratic is a frontier AI research and infrastructure company building a new class of LLMs. We're pushing foundational change at the model architecture level, enabling large-context and multi-modal inference that scales efficiently where transformers can't.
See moreDiscover more
Similar signals
Similar public activity from other companies.
Research & Knowledge
alphaXiv
alphaXiv published a paper introducing Declarative Attention for long-context language models that explicitly switches between global, focused, and local attention during reasoning to cut attended tokens by 52% on Gemma-4-31B and 31% on Qwen-3.6-27B.
Research & Knowledge
NEC Laboratories America, Inc.
NEC Laboratories America, Inc. published a paper titled Stop When Enough: Adaptive Early-Stopping for Chain-of-Thought Reasoning on July 8, 2026.
Research & Knowledge
OpenAI
OpenAI published a graph comparing GPT-6 Astra with a new model trained internally, which performs substantially better at comparable compute budgets and solves almost half of the problems at the highest budget shown.
Research & Knowledge
Applied Compute
Applied Compute developed a 35B open-weight model trained to search a precomputed index answering repo search questions at 100x lower cost than a frontier model.
Research & Knowledge
Together AI