Contact
🔒 Contacts are masked. Get full emails & phones. Unlock via export ↗
Category & trade
Primary
computing · embed-ft 0.38
Tree path
Technology & Computing › Computing
Group (tier-1)
Also (top-3)
Tech stack
AI readiness
AI-Readiness score
63 / 100 · ready · see similar
AI training policy
allowed
AI-bot protection
AI files
Evidence
llms.txt
Compliance (GEO / GDPR)
TLD
Overview
Title
JANGQ — Smaller Models, Bigger MLX Wins
Description
Curated JANGQ benchmarks showing only decisive wins where JANG is smaller than the MLX baseline: MiniMax +47.5 MMLU points, Qwen3.5-122B +33 points, and coherent smaller-bit runs on Mistral, Qwen2.5, SmolLM2, and TinyLlama.
Final URL
Language
en (html)
Scanned at
2026-07-07 11:01:47