NaceAI's tiny Drex tops Decision Index, edges out Jev and offers 250M free tokens
Hesamation · x · 2026-09-25
NaceAI released Drex, a decision-focused small model under 6B parameters: facts in, a probability for every option out in one pass, no text parsing. It scores 51.73 on Decision Index 0.2, narrowly beating Jev 1.13.0 (51.67) to rank first among 50 entries, while other top-ten models with published sizes range from 12B to 36B.
Drex is a diffusion model trained with RL from adjusted feedback, with 136ms response time (1.5x faster than Jev), aimed at agent routing, tool selection, reranking and guardrails. Demos show it making real-time decisions in StarCraft II, Doom, Lemmings and more. The first 10,000 builders get 250M free tokens.
More from Models
- Micius-Flash: 27B Multimodal Model for Hardware and Semiconductor Agents — zhiman-ai · 2026-11-11
- Claude IP risk-check site goes viral as users audit their IPs ahead of new rules — vista8 · 2026-11-11
- Baidu says RSI auto-optimized ERNIE inference, 6.24x decode speedup on Kunxin P800 — xiaohu · 2026-11-11
- Baidu's ECHO: multi-token prediction reusing the last layer as a drafter — xiaohu · 2026-11-11
- ERNIE 5.5 Flash Preview: 280B-param multi-router MoE activating only 14B per token — xiaohu · 2026-11-11
- ERNIE 5.5 recreates the Pelican test: 117-second animation drawn frame-by-frame in Python — xiaohu · 2026-11-11