WinoGrande

Models scored

137

Best score

0.89

Metric

Accuracy

WinoGrande is an AI evaluation tracked by Epoch AI, with 137 scored model versions on record. The scores here are reported by third parties rather than produced by Epoch AI running the evaluation itself.

Scored models span November 2019 to December 2024. The best score recorded is 0.89, measured as Accuracy. Models come from OpenAI, Google Research, DeepMind, Meta AI, Mistral AI, Anthropic, MosaicML, Technology Innovation Institute and others.

Full record

Score metric: Accuracy
Models scored: 137
Best score recorded: 0.89
Earliest model scored: Nov 5, 2019
Most recent model scored: Dec 26, 2024
Third-party reported: Yes

Leaderboard

01	Llama-3.1-405B	Meta AI	0.89
02	claude-3-opus-20240229	Anthropic	0.89
03	gpt-4-0314	OpenAI	0.88
04	gpt-4-32k-0314	OpenAI	0.88
05	falcon-180B	Technology Innovation Institute	0.87
06	DeepSeek-V2	DeepSeek	0.86
07	DeepSeek-V3	DeepSeek	0.85
08	PaLM 540B	Google Research	0.85
09	DeepSeek-Coder-V2-Base	DeepSeek	0.84
10	Meta-Llama-3-70B	Meta AI	0.83
11	PaLM 2-L		0.83
12	Qwen2.5-72B	Alibaba	0.82
13	Llama-3.1-405B-Instruct	Meta AI	0.82
14	text-davinci-002	OpenAI	0.82
15	gpt-3.5-turbo-0613	OpenAI	0.82
16	Phi-3-small-8k-instruct	Microsoft	0.81
17	Phi-3-medium-128k-instruct	Microsoft	0.81
18	Qwen2.5-Coder-32B	Alibaba	0.81
19	Llama-2-70b-hf	Meta AI	0.8
20	PaLM 2-M		0.79
21	GLaM (MoE)	Google	0.79
22	gemma-7b	Google DeepMind	0.79
23	Megatron-Turing NLG 530B	Microsoft,NVIDIA	0.79
24	falcon-11b	Technology Innovation Institute	0.78
25	Nemotron-4 15B	NVIDIA	0.78
26	PaLM 2-S		0.78
27	text-davinci-001	OpenAI	0.78
28	Mixtral-8x7B-v0.1	Mistral AI	0.77
29	LLaMA-65B	Meta AI	0.77
30	PaLM 62B		0.77
31	falcon-40b	Technology Innovation Institute	0.77
32	Qwen2.5-Coder-14B		0.77
33	Llama-2-34b	Meta AI	0.77
34	LLaMA-33B	Meta AI	0.76
35	Meta-Llama-3-8B	Meta AI	0.76
36	Mistral-7B-v0.1	Mistral AI	0.75
37	claude-3-sonnet-20240229	Anthropic	0.75
38	Chinchilla (70B)	DeepMind	0.75
39	claude-3-haiku-20240307	Anthropic	0.74
40	phi-1_5	Microsoft	0.73

More benchmarks

Epoch Capabilities Index MMLU Terminal-Bench GPQA Diamond OTIS Mock AIME 2024-2025 GSM8K ARC-AGI ARC-AGI-2

SourceEpoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/benchmarks. Licensed under CC BY 4.0.

← All benchmarks