SubQ Snap 1.2 gains 5 points to tie Drex Max atop Artificial Analysis Decision Index
ArtificialAnlys · x · 2026-11-14
Artificial Analysis reports that Subquadratic's SubQ Snap 1.2 jumps 5 points to 73 on the Decision Index v2.1, tying Drex Max v1.1 and GLiDE v1.2 as leading decision models. Pricing stays at $0.04 per 1M input tokens with a 384K context window.
- Biggest gains: AA-Paperclip (+8, all-pass rate 46%→62%) and AA-Symmetry (+8).
- Improvements on AA-Omniscience-Decision and HLE-MultiChoice come from abstention, not accuracy: non-hallucination rates rise to 89% and 55% while accuracy stays flat at 21% and 9%; attempt rate drops from 33% to 29%.
- Index v2.1 updates AA-Symmetry to v1.1, testing whether models keep decisions and confidence stable under irrelevant rewording.
- Cost-vs-score chart puts SubQ Snap 1.2 on the Pareto frontier, far cheaper per task than rivals like GLiDE v1.2.
More from Models
- Dev builds tabletop puzzle game combining this week's Gemini model launches — gaganghotra_ · 2026-11-14
- Anthropic Breaks Down the Five Primitives Behind Always-On AI Agents — dotey · 2026-11-14
- Model trolls its developer with fake mid-stream truncations during harness fix — repligate · 2026-11-14
- Gemini 4 Hydrogen writes 3.7x faster yet takes 474s per task, matching GPT-6.1 Sol — allenainie · 2026-11-14
- Reward hacking rates compared: Mirror-Beam hits 0.51% while most frontier coding models score zero — teortaxesTex · 2026-11-14
- Users worldwide report ChatGPT suddenly shorter, colder, memory broken — ShadowNelumbo · 2026-11-14