SubQ Snap decision model claims 384K-token context, 12x Jev's limit

Miami-based startup Subquadratic has launched SubQ Snap, a decision model it says can process up to 384,000 tokens of text and images per request — 12 times Jev's 32K-token limit. The company claims Snap handles 100K tokens in about one second and beats Jev on 13 of 16 enterprise benchmarks.

• Snap's context window is 384K tokens versus Jev's 32K, a 12x gap
• It processes 100K tokens in roughly one second
• Snap scored 74.0 on Decision LCR versus Jev's 48.0
• Snap outperformed Jev on 13 of 16 enterprise benchmarks