legal benchmark
SuperGPQA Leaderboard
SuperGPQA evaluates large language models on 25,957 questions spanning 285 graduate-level academic disciplines across 13 broad disciplinary areas.
Updated Sep 7, 2026
Models34
Model coverage34
MetricScore
EvidenceB
SuperGPQA Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score73.60% | Percentile100.00% | Participants34 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score71.60% | Percentile96.97% | Participants34 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score71.40% | Percentile93.94% | Participants34 | EvidenceC | Evaluated |
| Rank04 | ModelBY | Score70.80% | Percentile90.91% | Participants34 | EvidenceC | Evaluated |
| Rank05 | ModelAC | Score70.40% | Percentile87.88% | Participants34 | EvidenceC | Evaluated |
| Rank06 | ModelBY | Score67.40% | Percentile84.85% | Participants34 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score67.10% | Percentile81.82% | Participants34 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score66.00% | Percentile78.79% | Participants34 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score65.60% | Percentile75.76% | Participants34 | EvidenceC | Evaluated |
| Rank10 | ModelAC | Score65.10% | Percentile72.73% | Participants34 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score64.90% | Percentile69.70% | Participants34 | EvidenceC | Evaluated |
| Rank12 | ModelAC | Score64.70% | Percentile66.67% | Participants34 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score64.30% | Percentile63.64% | Participants34 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score63.40% | Percentile60.61% | Participants34 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score62.60% | Percentile57.58% | Participants34 | EvidenceC | Evaluated |
| Rank16 | ModelAC | Score60.80% | Percentile54.55% | Participants34 | EvidenceC | Evaluated |
| Rank17 | ModelAC | Score60.40% | Percentile51.52% | Participants34 | EvidenceC | Evaluated |
| Rank18 | ModelAC | Score59.00% | Percentile48.48% | Participants34 | EvidenceC | Evaluated |
| Rank19 | ModelAC | Score58.80% | Percentile45.45% | Participants34 | EvidenceC | Evaluated |
| Rank20 | ModelAC | Score58.20% | Percentile42.42% | Participants34 | EvidenceC | Evaluated |
| Rank21 | ModelMA | Score57.20% | Percentile39.39% | Participants34 | EvidenceC | Evaluated |
| Rank22 | ModelMA | Score57.20% | Percentile36.36% | Participants34 | EvidenceC | Evaluated |
| Rank23 | ModelAC | Score56.40% | Percentile33.33% | Participants34 | EvidenceC | Evaluated |
| Rank24 | ModelAC | Score54.60% | Percentile30.30% | Participants34 | EvidenceC | Evaluated |
| Rank25 | ModelAC | Score53.10% | Percentile27.27% | Participants34 | EvidenceC | Evaluated |
| Rank26 | ModelAC | Score52.90% | Percentile24.24% | Participants34 | EvidenceC | Evaluated |
| Rank27 | ModelAC | Score51.20% | Percentile21.21% | Participants34 | EvidenceC | Evaluated |
| Rank28 | ModelAC | Score46.80% | Percentile18.18% | Participants34 | EvidenceC | Evaluated |
| Rank29 | ModelMA | Score44.70% | Percentile15.15% | Participants34 | EvidenceC | Evaluated |
| Rank30 | ModelAC | Score44.50% | Percentile12.12% | Participants34 | EvidenceC | Evaluated |
SuperGPQA Highlights
The leading models and scores on this benchmark.
SuperGPQA Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for SuperGPQA
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is SuperGPQA?
What SuperGPQA measures and how its scores work.
SuperGPQA is a benchmark containing questions across graduate-level academic disciplines, created using a Human-LLM collaborative filtering mechanism with over 80 expert annotators.
It measures graduate-level knowledge and reasoning capabilities across fields including Engineering, Medicine, Science, Law, light industry, agriculture, and service-oriented domains.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- SuperGPQA
- Modality
- text
- Primary category
- legal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about SuperGPQA.
Which model scores highest on SuperGPQA?
Qwen3.7 Max is currently ranked first with 73.60%.
What are the top three models on SuperGPQA?
The current leaders are Qwen3.7 Max (73.60%), Qwen3.6 Plus (71.60%), and Qwen3.7-Plus (71.40%).
Which SuperGPQA model has the lowest official input price?
Qwen3.6-35B-A3B has the lowest matched official input price at $0.25 input / $1.5 output per 1M tokens.
Which models are fastest among SuperGPQA results?
The fastest matched records are Qwen3-235B-A22B-Instruct-2507 (21.74 tok/s via DeepInfra), Qwen3 235B A22B (21.74 tok/s via DeepInfra), and Qwen3 VL 4B Thinking (8.77 tok/s via DeepInfra).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.