Bidraft, an AI startup from South Korea, has analyzed 63 official benchmarks from Hugging Face, revealing that its Darwin AI models ranked first in 12 categories. This achievement places Bidraft ahead of 140 participating organizations, with Moonshot AI and Z.ai following in second and third place respectively. The analysis highlighted Bidraft's strong performance across diverse fields including science, mathematics, document recognition, and law, with its Darwin-180B model also topping a SuperCLUE evaluation in China. AI
IMPACT Demonstrates the broad capabilities of Bidraft's Darwin models across various domains, potentially influencing future AI development and competition.
RANK_REASON Analysis of benchmark results from an AI platform. [lever_c_demoted from research: ic=1 ai=1.0]
- AIME 2026
- Alibaba Qwen
- Ant Group
- Bidraft
- DeepSeek
- ExtractBench
- GPQA Diamond
- HMMT 2026
- Hugging Face
- KDLAI
- Kim Min-sik
- LEXam-hard
- MDPBench
- Minimax
- Moonshot AI
- Thinking machines
- Xiaomi
- Z.ai
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →