← Back to Tools

SubQ

hub AI Models & Platforms Freemium

First frontier model with Subquadratic Sparse Attention (SSA). 12M token context, 52x faster than FlashAttention at 1M tokens, costs under 5% of Claude Opus.

LLMLong ContextSSASparse Attention12M TokensSubquadratic
Visit Website arrow_outward

Enjoying this?

Subscribe to get new solo-founder resources when they are actually published.

No spam. Unsubscribe anytime.