The AI arena is free today

Open Superagent
AI OrganizationMaintains 5 benchmarks

Tsinghua University: AI benchmarks & research

Tsinghua University research groups (THUDM, THUNLP and others). Maintains MedXpertQA, OlympiadBench and OVO-Bench, and co-maintains LongBench v2 with Zhipu AI.

Maintained by Tsinghua University

5 benchmarks

In one line

Tsinghua University maintains LongBench v2 (counts toward the LLM Stats Score, 19 models scored), MedXpertQA (counts toward the LLM Stats Score, 12 models scored), MedXpertQA-MM (counts toward the LLM Stats Score, 1 model scored), OlympiadBench (counts toward the LLM Stats Score, 1 model scored) and OVOBench (counts toward the LLM Stats Score, 4 models scored). It ships no models of its own on LLM Stats.

FAQ

Common questions about Tsinghua University.

What is Tsinghua University?

Tsinghua University research groups (THUDM, THUNLP and others). Maintains MedXpertQA, OlympiadBench and OVO-Bench, and co-maintains LongBench v2 with Zhipu AI. On LLM Stats it maintains 5 benchmarks, led by LongBench v2, MedXpertQA and OVOBench. It ships no models of its own; each score on its benchmarks carries its own provenance label on the model page.

Which Tsinghua University benchmarks count toward the LLM Stats Score?

LongBench v2, MedXpertQA, OVOBench, MedXpertQA-MM and OlympiadBench count toward the LLM Stats Score. Only benchmarks admitted to the core set move the main leaderboard; the rest are shown for transparency.

Which models lead Tsinghua University's benchmarks?

Beam 1.0 leads LongBench v2 at 67.5% (19 models scored), Muse Spark leads MedXpertQA at 78.4% (12 models scored) and Seed 2.1 Pro leads OVOBench at 80.7% (4 models scored). Each benchmark page lists the full ranking and how the results were produced.

Where are Tsinghua University's official benchmark results?

Tsinghua University's official site is https://www.tsinghua.edu.cn. Each benchmark page on LLM Stats links to that benchmark's own paper or leaderboard, and shows the scores next to every other benchmark so models can be compared on one page.