Skip to content

MLE-Bench Lite

A lightweight machine-learning competition benchmark that measures whether models can iteratively train, evaluate, and improve ML systems in low-resource settings.

2 models

Higher is better.

  1. 1Atria Dawn PreviewShanghai Artificial Intelligence Laboratory86.2
  2. 2MiniMax M2.7MiniMax66.6
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Autonomous iterative ML optimization
Tasks
Low-resource ML competitions
Authors
MiniMax (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.