reasoning benchmark
CommonSenseQA Leaderboard
CommonSenseQA is a multiple-choice question answering dataset that requires different types of commonsense knowledge to predict correct answers. It contains 12,102 questions with one correct answer and four distractors, designed to test semantic reasoning and conceptual relationships. Questions are created based on ConceptNet concepts and require prior world knowledge for accurate reasoning.
Updated Aug 17, 2026
Models1
Model coverage1
MetricScore
EvidenceB
CommonSenseQA Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelMA | Score70.4% | Percentile100.0% | Participants1 | EvidenceC | Evaluated |
CommonSenseQA Highlights
The leading models and scores on this benchmark.
The Top AI Models for CommonSenseQA
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis commonsenseqa AI model leaderboard uses descending score in the benchmark's original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
- 01MAMistral NeMo InstructMistral AI
- Score
- 70.4%
- Price
- $0.15 input / $0.15 output per 1M tokens
- Speed
- Up to 42 tok/s via Google
Strengths
- Ranks #1 of 1 compared models
- 100th percentile on this benchmark
- C evidence result
Considerations
- This result measures CommonSenseQA, not total model capability
Selection summary
Best AI Models for CommonSenseQA
Mistral NeMo Instruct currently leads CommonSenseQA with 70.4%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.
What is CommonSenseQA?
What CommonSenseQA measures and how its scores work.
CommonSenseQA is a multiple-choice question answering dataset that requires different types of commonsense knowledge to predict correct answers. It contains 12,102 questions with one correct answer and four distractors, designed to test semantic reasoning and conceptual relationships. Questions are created based on ConceptNet concepts and require prior world knowledge for accurate reasoning.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- CommonSenseQA
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- commonsenseqa|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about CommonSenseQA.
Which model scores highest on CommonSenseQA?
Mistral NeMo Instruct is currently ranked first with 70.4%.
What does CommonSenseQA measure?
CommonSenseQA is a multiple-choice question answering dataset that requires different types of commonsense knowledge to predict correct answers. It contains 12,102 questions with one correct answer and four distractors, designed to test semantic reasoning and conceptual relationships. Questions are created based on ConceptNet concepts and require prior world knowledge for accurate reasoning.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.