Artificial Analysis Decision Index Snapshot Maps 23 Competing Decision Models
An independent five-evaluation snapshot across 23 decision models places Liquid d1-ultra first at 65.2, Jev 1.14 at 57.1, and OpenAI’s GPT-6 Luna Decisions at 48.5.
As vendor-published benchmarks multiplied across October, independent tracker Artificial Analysis released a November 5 snapshot of its Decision Index evaluating 23 hosted and open-weight decision models across five shared tasks, including hallucination resistance and tool-calling decisions.
Where Jev 1.14 stands in the crowded field
In the November 5 ranking, Liquid AI’s d1-ultra led the board at 65.2—priced at the same $0.042 per million input tokens as TypeSafe AI’s flagship model. Jev 1.14 placed 10th overall at 57.1, comfortably ahead of OpenAI’s GPT-6 Luna Decisions API at 48.5 (21st place).
The snapshot also underscored how quickly specialized entrants have grown from a single product category into a 23-model market spanning Liquid AI, Perplexity, Cloudflare, Microsoft, TypeSafe, OpenAI, Upstage, Inception, Baseten, and open-source teams.
Reading public indexes alongside calibration and latency
Independent evaluators stressed that composite index scores are only one axis of a production decision. Jev 1.14 documents a 64K-token request budget (expandable to 262K via header) and was specifically trained with Reinforcement Learning for Calibrated Decisions (RLCD) so confidence values correlate closely with real-world accuracy. For teams routing live traffic, latency variance, calibration error, and performance on domain-specific schemas often outweigh a few points on a generalized leaderboard.