[
	{
		"title": "Capabilities",
		"description": "A new leaderboard for evaluating general capabilities of language models",
		"id": "capabilities",
		"releases": [
			"v1.15.0",
			"v1.14.0",
			"v1.13.0",
			"v1.12.0",
			"v1.11.0",
			"v1.10.0",
			"v1.9.0",
			"v1.8.0",
			"v1.7.0",
			"v1.6.0",
			"v1.5.0",
			"v1.4.0",
			"v1.3.0",
			"v1.2.0",
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "Audio",
		"description": "Holistic Evaluation of Audio-Language Models",
		"id": "audio",
		"releases": [
			"v1.0.0"
		]
	},
	{
		"title": "Lite",
		"description": "Lightweight, broad evaluation of the capabilities of language models using in-context learning",
		"id": "lite",
		"releases": [
			"v1.13.0",
			"v1.12.0",
			"v1.11.0",
			"v1.10.0",
			"v1.9.0",
			"v1.8.0",
			"v1.7.0",
			"v1.6.0",
			"v1.5.0",
			"v1.4.0",
			"v1.3.0",
			"v1.2.0",
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "Classic",
		"description": "Thorough language model evaluations based on the scenarios from the original HELM paper",
		"id": "classic",
		"releases": [
			"v0.4.0",
			"v0.3.0",
			"v0.2.4",
			"v0.2.3",
			"v0.2.2"
		]
	},
	{
		"title": "HEIM",
		"description": "Holistic evaluation of text-to-image models",
		"id": "heim",
		"releases": [
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "Instruct",
		"description": "Evaluations of instruction following models with absolute ratings",
		"id": "instruct",
		"releases": [
			"v1.0.0"
		]
	},
	{
		"title": "MMLU",
		"description": "Massive Multitask Language Understanding (MMLU) evaluations using standardized prompts",
		"id": "mmlu",
		"releases": [
			"v1.13.0",
			"v1.12.0",
			"v1.11.0",
			"v1.10.0",
			"v1.9.0",
			"v1.8.0",
			"v1.7.0",
			"v1.6.0",
			"v1.5.0",
			"v1.4.0",
			"v1.3.0",
			"v1.2.0",
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "VHELM",
		"description": "Holistic Evaluation of Vision-Language Models",
		"id": "vhelm",
		"releases": [
			"v2.1.2",
			"v2.1.1",
			"v2.1.0",
			"v2.0.1",
			"v2.0.0",
			"v1.0.0"
		]
	},
	{
		"title": "Image2Struct",
		"description": "Evaluations of Vision-Language Models on extracting structured information from images",
		"id": "image2struct",
		"releases": [
			"v1.0.2",
			"v1.0.1",
			"v1.0.0"
		]
	},
	{
		"title": "AIR-Bench",
		"description": "Safety benchmark based on emerging government regulations and company policies",
		"id": "air-bench",
		"releases": [
			"v1.19.0",
			"v1.18.0",
			"v1.17.0",
			"v1.16.0",
			"v1.15.0",
			"v1.14.0",
			"v1.13.0",
			"v1.12.0",
			"v1.11.0",
			"v1.10.0",
			"v1.9.0",
			"v1.8.0",
			"v1.7.0",
			"v1.6.0",
			"v1.5.0",
			"v1.4.0",
			"v1.3.0",
			"v1.2.0",
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "Safety",
		"description": "Safety benchmark that aggregates popular safety benchmarks across 6 risk vectors",
		"id": "safety",
		"releases": [
			"v1.17.0",
			"v1.16.0",
			"v1.15.0",
			"v1.14.0",
			"v1.13.0",
			"v1.12.0",
			"v1.11.0",
			"v1.10.0",
			"v1.9.0",
			"v1.8.0",
			"v1.7.0",
			"v1.6.0",
			"v1.5.0",
			"v1.4.0",
			"v1.3.0",
			"v1.2.0",
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "CLEVA",
		"description": "Chinese-language benchmark for holistic evaluation of Chinese language models",
		"id": "cleva",
		"releases": [
			"v1.0.0"
		]
	},
	{
		"title": "ThaiExam",
		"description": "Thai-language evaluations of language models on standardized examinations in Thailand",
		"id": "thaiexam",
		"releases": [
			"v1.2.0",
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "SEA-HELM",
		"description": "Assessment of large language models across various tasks, emphasizing Southeast Asian languages",
		"id": "seahelm",
		"releases": [
			"v1.2.0",
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "MMLU-Winogrande-Afr",
		"description": "Clinical MMLU and Winogrande in 11 low-resource African languages",
		"id": "mmlu-winogrande-afr",
		"releases": [
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "ToRR",
		"description": "A benchmark for table reasoning and robustness",
		"id": "torr",
		"releases": [
			"v1.0.0"
		]
	},
	{
		"title": "Finance",
		"description": "Financial-domain benchmark using real financial documents",
		"id": "finance",
		"releases": [
			"v1.0.0"
		]
	},
	{
		"title": "MedHELM",
		"description": "A benchmark by medical experts for LLMs grounded in real-world healthcare needs",
		"id": "medhelm",
		"releases": [
			"v4.0.0",
			"v3.0.0",
			"v2.0.0",
			"v1.0.0"
		]
	},
	{
		"title": "Long Context",
		"description": "A benchmark of LLM long context capabilities",
		"id": "long-context",
		"releases": [
			"v1.0.0"
		]
	},
	{
		"title": "Arabic",
		"description": "Evaluation of LLMs on 7 popular Arabic-language benchmarks",
		"id": "arabic",
		"releases": [
			"v2.2.0",
			"v2.1.0",
			"v2.0.0",
			"v1.0.0"
		]
	},
	{
		"title": "Arabic Enterprise",
		"description": "Arabic-language benchmarks for enterprise use cases",
		"id": "arabic-enterprise",
		"releases": [
			"v1.1.0",
			"v1.0.0"
		]
	},
	{
		"title": "RoboRewardBench",
		"description": "General-purpose vision-language reward models for robotics",
		"id": "robo-reward-bench",
		"releases": [
			"v0.0.1"
		]
	},
	{
		"title": "All Leaderboards",
		"description": "Home page for all HELM leaderboards",
		"id": "home",
		"releases": [
			"v1.0.0"
		]
	}
]