Menu

How a 180B Open Model Outscored a 2.8T Model in Math and Science Reasoning: Strength Merging Plus Model-Level RSI
📰

How a 180B Open Model Outscored a 2.8T Model in Math and Science Reasoning: Strength Merging Plus Model-Level RSI

DEV Community: opensource·AI OpenFree·4 days ago
#9z7x4JCJ
#devtoopensource#devto#dev#model#darwin
Reading 0:00
15s threshold

TL;DR

Darwin-180B-RSI , an open-weight model from Korea's VIDRAFT, scored 73.68 in mathematical and scientific reasoning on SuperCLUE's November 2026 evaluation. That is first among open-weight models and second overall, directly behind OpenAI's GPT-6.1-Sol (76.32). The largest model in the same evaluation, Moonshot AI's Kimi-K3.1-Pro at 2.8 trillion parameters, scored 63.16. Darwin has 180 billion parameters, about one sixteenth of that size, and finished more than 10 points ahead.

This post explains the two techniques behind the result and what the size difference means for serving cost.

The full ranking

Why SuperCLUE counts as third-party verification

SuperCLUE does not accept scores submitted by model developers. The evaluator calls each model directly and grades it on a private test set that changes every month.…

Continue reading — create a free account

Join HashtagPLUS to read full articles, follow hashtags, vote, and join the conversation.