Commercial training data from specialist providers. Listings show what each source advertises; licensing is confirmed with the provider before any agreement.
A single-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Egyptian Arabic.
A multilingual AI model evaluation dataset built by having expert reviewers re-evaluate and verify automatically scored response quality according to defined evaluation…
A multilingual instruction-following QA text dataset built for tasks that require generating answers based on referenced documents.
A multilingual STEM preference evaluation dataset built from STEM-domain questions and three AI-generated answers, with STEM majors providing scoring, feedback, and…
A multi-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Bengali.
A single-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Bengali.
An English safety multi-turn chat text dataset built by professional annotators for safety alignment training based on defined safety criteria.
A Korean-based Chat Arena preference evaluation multi turn dataset built by collecting human preferences for AI model responses.
A Korean-based Chat Arena preference evaluation single-turn dataset built by collecting human preferences for AI model responses.
A multi-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Gulf Arabic.
A single-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Gulf Arabic.
A multi-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Hindi.
A single-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Hindi.
A multi-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Indonesian.
A single-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Indonesian.
A multi-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Japanese.
A single-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Japanese.
A Korean document summarization text dataset built by converting Korean news articles into markdown-based outputs such as key trends, summary reports, and document…
A Korean RAG instruction text dataset built from documents with user queries, query type labels, answers, and supporting evidence spans written in Korean.
Real-time prediction probabilities and trends analyzing the future outcomes of global and domestic political, electoral, economic, and cultural issues through crowd…
A multi-turn conversational dataset designed to evaluate model response capabilities against major safety risk categories and attack patterns.
A multi-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Thai.
A single-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Thai.