New Model

Contribute to New Model

In today’s rapidly evolving landscape of multimodal AI, countless models are emerging with varying capabilities and levels of generality. At General-Bench, our goal is to provide a comprehensive, and reliable evaluation framework to assess the true intelligence level of these models across a broad spectrum of tasks and modalities.

We strongly encourage users and researchers to submit new models to our platform for evaluation. This helps ensure consistent benchmarking and promotes transparency in model capability reporting.

To accommodate different user goals and evaluation budgets, we offer multiple dataset types and evaluation tracks. These are designed to meet the needs of both large-scale generalist models and smaller, task-specific systems.

To get started with evaluating your model, please follow the steps outlined in the tutorial below:

• Overview
• Prepare Files
• Prepare Models
• Run Evaluation
• Submit Results