NVDA 125.53 ▼14.01%GOOGL 265.29 ▼2.70%MSFT 484.52 ▼2.77%AMD 424.39 ▼13.25%INTC 51.28 ▼7.97%TSMC 318.71 ▼9.48%AMZN 246.82 ▼4.10%META 560.84 ▼5.93%AAPL 342.56 ▼1.90%PLTR 143.49 ▼6.63%
Markets at last close

Models

Kimi K3.1 Pro leads modelgrep’s open-weight rankings

·1 min read

Kimi K3.1 Pro leads modelgrep’s November 2026 ranking of open-weight language models with an intelligence score of 46.5. The leaderboard covers 25 models that are downloadable from Hugging Face and self-hostable, positioning them as alternatives to proprietary frontier systems.

Hy4 follows with 34.6, while DeepSeek V4.1 Flash 1017 scores 32.6. Generation speeds differ across the leaders: Kimi K3.1 Pro is listed at 29 t/s, compared with 95 t/s for Hy4 and 160 t/s for DeepSeek V4.1 Flash 1017. Hy4 is the closest alternative to Kimi on the intelligence metric used for the ranking.

Other highly placed models include GLM 5.3, with an intelligence score of 31.7, and TInkling 2 Small, at 30.6. The listings also show pricing and capability labels, including reasoning, tool use and JSON support, providing additional comparison points beyond the intelligence scores that determine the ordering.

Originally reported by modelgrep.comRead the source →
Related coverage