OmniGAIA

reasoning

Multilingual: no

OmniGAIA evaluates multimodal perception and reasoning in agentic contexts, testing a model's ability to process diverse inputs and perform complex multi-step reasoning tasks.

Methodology

Imported from llm-stats public benchmark metadata. Modality: multimodal. Max score: 1. Categories: agents, multimodal, reasoning. Language: en. Verified by llm-stats: no.