Reasoning Benchmark

WinoGrande

WinoGrande - large-scale commonsense reasoning with Winograd schema-like problems. 5-shot.

Category

Reasoning

Score type

percent

Scale

100%

higher is better

Models tested

6

What is WinoGrande?

WinoGrande is a benchmark for general reasoning and comprehension. The official description says: "WinoGrande - large-scale commonsense reasoning with Winograd schema-like problems". Below we describe in simple words what benchmarks of this type generally examine.

What does it test?

WinoGrande measures general reasoning and comprehension. In detail, it checks:

  • drawing the right conclusion from a passage or scenario
  • commonsense judgement about everyday situations and words
  • counting, comparing and reasoning over several facts at once
  • avoiding shortcuts that memorise answers instead of reasoning

How to read the score

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better.

Official methodology note: WinoGrande - large-scale commonsense reasoning with Winograd schema-like problems. 5-shot.

Frequently asked questions

What is the WinoGrande benchmark?

WinoGrande is a benchmark for general reasoning and comprehension. The official description says: 'WinoGrande - large-scale commonsense reasoning with Winograd schema-like problems'. Below we describe in simple words what benchmarks of this type generally examine.

What does WinoGrande test?

WinoGrande mainly measures general reasoning and comprehension. Concretely, it checks: drawing the right conclusion from a passage or scenario; commonsense judgement about everyday situations and words; counting, comparing and reasoning over several facts at once.

How is the WinoGrande score calculated?

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better. The declared maximum score of this benchmark is 100.0.

Where do I get WinoGrande data and scores?

The official source is https://arxiv.org/abs/1907.10641. The official benchmark page/reference is linked on this page. Every score on Open LLM Radar is also available as open data: download the scores JSON (http://open-llm-radar-api:8000/api/v1/benchmarks/winogrande/scores) or pull it directly from the Open LLM Radar API with any HTTP client.

Model scores 6 models tested

Reset
# Model Score Date Source Verified
1 Clef-Flash
Cloudflare
97.5%
— Report Verified
2 Clef
Cloudflare
93.5%
— Report Verified
3 Gemma 3 27B PT
Google DeepMind
78.8%
27.02.2025 Report Verified
4 Gemma 3 12B PT
Google DeepMind
74.3%
27.02.2025 Report Verified
5 Gemma 3 4B PT
Google DeepMind
64.7%
27.02.2025 Report Verified
6 Gemma 3 1B PT
Google DeepMind
58.2%
27.02.2025 Report Verified

What now?

You've seen how models score on WinoGrande — here's what to do next.