DMind Benchmark Included in KDD 2026, First Evaluation Benchmark for Digital Assets × AI Large Models

The DMind Benchmark, launched by the Minara team, was released in 2025 and its paper has recently been officially accepted by the Datasets & Benchmarks Track of KDD 2026, becoming the first large model evaluation benchmark in the intersection of digital assets and AI to pass peer review and enter an international top conference. KDD is a premier international conference in the field of data mining and machine learning, with an acceptance rate of approximately 29% for this track. The DMind Benchmark measures the multi-step reasoning capabilities of models in real-world scenarios of digital assets, consisting of 3,154 objective questions and 389 open-ended tasks, covering nine subfields including DeFi, Tokenomics, and contract security. It has systematically evaluated 31 cutting-edge large models such as GPT, Claude, and Gemini. The dataset has topped the Trending section on Hugging Face, with over 13,000 downloads. A re-evaluation in July 2026 showed that the progress of six latest flagship models on this benchmark has not bridged the depth gap in the field.