DigestAI news desk

Cut through the AI noise.

Research7 min read

NVIDIA releases Kumo Tabular model for tabular prediction with open weights

NVIDIA has launched Kumo Tabular, an open-source foundation model for tabular data prediction, available on Hugging Face. The model handles classification and regression tasks without training, tuning, or feature engineering, using in-context learning on labeled tables. It comes in three sizes (28M to 215M parameters) and is pretrained exclusively on artificial data generated via procedural…

1 source primary source

Key points

  • Kumo Tabular predicts labels in a single forward pass with no training, supporting classification and regression on tabular data
  • Ranks first on four benchmarks (TabArena, BeyondArena, TALENT, ScoringBench) and runs 17x faster than LimiX-2 on NVIDIA GPUs
  • Pretrained on 35–137 million artificial tables, with weights open-sourced under OpenMDW-1.1 for commercial use

Kumo Tabular achieves state-of-the-art performance on four benchmarks—TabArena, BeyondArena, TALENT, and ScoringBench—while running up to 17x faster than competing models like LimiX-2 on an RTX 6000 Pro. The model supports up to 10 classes natively and includes built-in preprocessing for numerical/categorical data. NVIDIA provides an open-source library for inference, with weights hosted on Hugging Face under the OpenMDW-1.1 license, allowing commercial use. The company emphasizes validation on held-out data due to potential accuracy degradation in edge cases.

Model page: Kumo Tabular →

Full story from Hugging Face primary sourceOpen source ↗

NVIDIA Kumo Tabular Sets a New Accuracy-Efficiency Frontier for Tabular Prediction

Hugging Face · 29 September 2026

Loading the full article…

This text was published by Hugging Face. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Topics · follow one to build your own front page
NVIDIAHugging FaceKumo TabularLimiX-2TabICLTabPFNDavid HolzmüllerVignesh Kothapalli

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.

Comments

via GitHub Discussions

More in Research

All →

Related stories