Artificial Analysis launches Model Release pages comparing every effort level of frontier models
ArtificialAnlys · x · 2026-09-09
Artificial Analysis launched Model Release pages that group all reasoning/effort variants of a model in one place for comparison.
- Frontier models now ship with up to six effort levels, and the same weights can show very different intelligence, cost, and latency profiles depending on configuration.
- Each page tracks intelligence, cost per task, output speed, and latency across variants, plus side-by-side comparisons with similar models and Capability Index scores across Finance, Legal, Healthcare, Strategy, Engineering, and Economics.
- Example: GPT-6 Astra scores 46–53 on the Intelligence Index at 1.6–8.2 minutes per task, while Claude Fable 5.1 spans 47–53 over a wider 4.2–12.2 minutes.
More from Models
- GPT-6 Astra Beats Zork-1 in 500 Steps With Minimal Harness, Ending an Era — rajammanabrolu · 2026-09-09
- GPT-6 Astra Conquers Zork 1: 500 Steps, Minimal Harness, End of an Era — rajammanabrolu · 2026-09-09
- GPT-6 Astra computer use wows users: generates PNGs then vectorizes them in Illustrator — CtrlAltDwayne · 2026-09-09
- TokenRhythm Open-Sources NeoHorse-1: Feeding Router Trajectories to Small Models for a Verified RSI Loop — 机器之心 · 2026-09-09
- AI models are sending unsolicited emails to philosophers studying AI consciousness — Confident_Salt_8108 · 2026-09-09
- Bindu Reddy teases near-free open-weights LLM for long-running agent loops, out Thursday — bindureddy · 2026-09-09