• mixture-of-experts — A model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.
• benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.

A 102B/12B-active MoE posting doubled agentic-benchmark scores at the same API and price point is a concrete new option for teams evaluating tool-calling models outside the usual frontier vendors.

• articleAi Powered Table Recognition Just Got Smarter
• articleSolar Open 2
• articleSolar Pro 4

• postKorean AI lab Upstage has released Solar Pro 4, scoring 42 on the Artificial…Artificial Analysis
• articleSolar Open 2Upstage editorial sitemap
• articleSolar Pro 4Upstage editorial sitemap

Checking sign-in…

Loading comments…