Subquadratic

Alexander Whedon Verified account @alex_whedon · Oct 12

Introducing SubQ Snap, the first long-context decision model for real enterprise work.

Snap is the only decision model or LLM where compute per token stays constant as context grows.

• 384K tokens of multimodal context, 12x larger than Jev.
• 61.7 on Decision Index v0.2.1 vs Jev’s 57.9, outperforming on 13 of 16 enterprise benchmarks.
•Processes 100K tokens and returns a decision in under one second.
•$0.05/MM input tokens. Output is free.

Available today to anyone.

The first 10,000 sign-ups get 100M free input tokens.

We’ll likely hit capacity fast and will add users as we add compute. Compute offers welcome 🙂.

Latency will be variable as we tune the platform to this new model and category of users.

Model speeds are 80ms for minimal context, 250ms for 32k, 660ms for 100k, and 2.5 seconds for 384k.

Visit: subq.ai

View quoted post on X

Baseten Verified account @baseten · Nov 2

Introducing SubQ Snap 1.1 on Baseten.

Built for long-context decision-making, SubQ Snap 1.1 by @subquadratic has a 384K context window.

It scores 65 on the Artificial Analysis Decision Index, tied for #1.

Explore SubQ Snap 1.1 on Baseten: baseten.co/library/subq-s…

View quoted post on X

Cerebral Valley Verified account @cerebral_valley · Nov 2

San Francisco 🌁

We’re hosting a hackathon for the @aidotengineer Code Summit, in partnership with @GoogleDeepMind, @MetaforDevs, @MongoDB, @Fetch_ai and @subquadratic.

🏆 $11,000 in cash prizes

📅 Nov 7 at Shack15

Apply below 👇

View quoted post on X