Fu-Benchmark is a meta-benchmark of the most influential evaluation suites used to measure and rank large language models. Use the search box to filter by name, topic, or model.

All modelsOpen Source modelsOpen Source / TrainingCommercial models All modalitiesTextImage All sizesXS (<2B)S (2–7B)M (7–30B)L (30–70B)XL (70B+) All countries / regions──────────AggregatesEuropean Union (EU)Europe (EU + UK, CH, NO, IS)EMEA (Europe + Middle East)──────────Individual Countries Benchmarks

#	Name	Topic	Description	Relevance	GitHub ★	Leader	Top %
1	A12D o	Diagram reasoning	A12D diagram reasoning benchmark for measuring multimodal understanding of annotated diagrams.	★★★★★		🇺🇸 Gemini 2.5 Pro	90.9%
2	AA-Index o	Multi-domain QA	Comprehensive QA index across diverse domains.	★★★★★		🇺🇸 Grok 4	73.2%
3	AA-LCR O	Long-context reasoning	A challenging benchmark measuring language models' ability to extract, reason about, and synthesize information from long-form documents ranging from 10k to 100k tokens.	★★★★★		🇺🇸 GPT-5	75.6%
4	ACP-Bench Bool O	Safety evaluation (boolean)	Safety and behavior evaluation with yes/no questions.	★★★★★		🇨🇳 Qwen3-32B	85.1%
5	ACP-Bench MCQ O	Safety evaluation (MCQ)	Safety and behavior evaluation with multiple-choice questions.	★★★★★		🇺🇸 Llama 3.3 70B	82.1%
6	AgentDojo O	Agent evaluation	Interactive evaluation suite for autonomous agents across tools and tasks.	★★★★★		🇺🇸 Claude 3.7 Sonnet	88.7%
7	AGIEval O	Exams	Academic and professional exam benchmark.	★★★★★		🇺🇸 Llama 3.1 405B Base	71.6%
8	AI2D O	Diagram understanding (VQA)	Visual question answering over science and diagram images.	★★★★★		🇺🇸 Molmo-72B	96.3%
9	Aider Code Editing o	Code editing	Measures interactive code editing quality within the Aider assistant workflow.	★★★★★		🇺🇸 Gemini 2.5 Pro	89.8%
10	Aider-Polyglot O	Code assistant eval	Aider polyglot coding leaderboard.	★★★★★		🇺🇸 GPT-5	88.0%
11	AIME 2024 O	Math (competition)	American Invitational Mathematics Examination 2024 problems.	★★★★★		🇺🇸 GPT-OSS 120B	96.6%
12	AIME 2025 O	Math (competition)	American Invitational Mathematics Examination 2025 problems.	★★★★★		🇺🇸 GPT-5 pro	96.7%
13	AIME25 o	Math (competition)	American Invitational Mathematics Examination 2025 benchmark (set AIME25).	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Thinking	83.1%
14	All-Angles Bench o	Spatial perception	All-Angles benchmark for spatial recognition and 3D perception.	★★★★★		🇨🇳 GLM-4.5V	56.9%
15	AlpacaEval O	Instruction following	Automatic eval using GPT-4 as a judge.	★★★★★	1849	🇨🇳 Qwen3-32B	64.2%
16	AlpacaEval 2.0 O	Instruction following	Updated AlpacaEval with improved prompts and judging.	★★★★★		🇨🇳 DeepSeek R1	87.6%
17	AMC-23 O	Math (competition)	American Mathematics Competition 2023 evaluation.	★★★★★		G QwQ-32B	98.5%
18	AndroidWorld o	Mobile agents	Benchmark for agents operating Android apps via UI automation.	★★★★★		🇨🇳 GLM-4.5V	57.0%
19	API-Bank o	Tool use	API-Bank tool-use benchmark.	★★★★★		🇺🇸 Llama 3.1 405B	92.0%
20	ARC-AGI-1 O	General reasoning	ARC-AGI Phase 1 aggregate accuracy.	★★★★★		🇺🇸 o3	75.7%
21	ARC-AGI-2 O	General reasoning	ARC-AGI Phase 2 aggregate accuracy.	★★★★★		🇺🇸 GPT-5 pro	18.3%
22	ARC Average O	Science QA (average)	Average accuracy across ARC-Easy and ARC-Challenge.	★★★★★		🇺🇸 SmolLM2 1.7B Pretrained	60.5%
23	ARC-Challenge O	Science QA	Hard subset of AI2 Reasoning Challenge; grade-school science.	★★★★★		🇺🇸 Llama 3.1 405B	96.9%
24	ARC-Challenge (DE) o	Science QA (German)	German translation of the ARC Challenge benchmark.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	0.7%
25	ARC-Easy O	Science QA	Easier subset of AI2 Reasoning Challenge.	★★★★★		🇺🇸 Gemma 3 PT 27B	89.0%
26	ARC-Easy (DE) o	Science QA (German)	German translation of the ARC Easy science QA benchmark.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	0.8%
27	Arena-Hard O	Chat ability	Hard prompts on Chatbot Arena.	★★★★★	920	🇫🇷 Mistral Medium 3	97.1%
28	Arena-Hard V2 O	Chat ability	Updated Arena-Hard v2 prompts on Chatbot Arena.	★★★★★	920	🇨🇳 Qwen3 MoE-2507	88.2%
29	ARKitScenes o	3D scene understanding	ARKitScenes benchmark for assessing 3D scene reconstruction and understanding from mixed reality captures.	★★★★★		🇨🇳 Qwen2.5-VL 72B Instruct	61.5%
30	ART Agent Red Teaming O	Agent robustness	Evaluation suite for adversarial red-teaming of autonomous AI agents.	★★★★★		🇺🇸 Claude Sonnet 4.5 (Thinking)	↓ 40.0%
31	ArtifactsBench O	Agentic coding	Artifacts-focused coding and tool-use benchmark evaluating generated code artifacts.	★★★★★		🇺🇸 GPT-5	72.5%
32	AstaBench O	Agent evaluation	Evaluates science agents across literature understanding, data analysis, planning, tool use, coding, and search.	★★★★★		🇺🇸 Claude Sonnet 4	53.0%
33	AttaQ O	Safety / jailbreak	Adversarial jailbreak suite measuring refusal robustness against targeted attack prompts.	★★★★★		G Granite 3.3 8B Instruct	88.5%
34	AutoCodeBench O	Autonomous coding	End-to-end autonomous coding benchmark with unit-test based execution across diverse repositories and tasks.	★★★★★		🇺🇸 Claude Opus 4 (Thinking)	52.4%
35	AutoCodeBench-Lite O	Autonomous coding	Lite version of AutoCodeBench focusing on smaller tasks with the same end-to-end, unit-test-based evaluation.	★★★★★		🇺🇸 Claude Opus 4	64.5%
36	BALROG O	Agent robustness	Benchmark for assessing LLM agents under adversarial and out-of-distribution tool-use scenarios.	★★★★★		🇺🇸 Grok 4	43.6%
37	BBH O	Multi-task reasoning	Hard subset of BIG-bench with diverse reasoning tasks.	★★★★★	510	🇨🇳 ERNIE 4.5 424B A47B	94.3%
38	BBQ O	Bias evaluation	Bias Benchmark for Question Answering evaluating social biases across contexts.	★★★★★		🇫🇷 Mixtral 8x 7B	56.0%
39	BFCL o	Code reasoning	Benchmark for functional code correctness and logic.	★★★★★		🇨🇳 Qwen3-4B	95.0%
40	BFCL Live v2 o	Finance QA	Financial compliance and literacy questions from the BFCL Live v2 benchmark.	★★★★★		🇺🇸 o1 Mini	81.0%
41	BFCL v3 O	Code reasoning	Benchmark for functional code correctness and logic (v3).	★★★★★		🇨🇳 GLM 4.5	77.8%
42	BIG-Bench o	Multi-task reasoning	BIG-bench overall performance (original).	★★★★★	3110	🇺🇸 Gemma 2 7B	55.1%
43	BIG-Bench Extra Hard o	Multi-task reasoning	Extra hard subset of BIG-bench tasks.	★★★★★		G Ling 1T	47.3%
44	BigCodeBench O	Code Generation	BigCodeBench evaluates large language models on practical code generation tasks with unit-test verification.	★★★★★		🇺🇸 GPT-4o-2024-05-13	56.1%
45	BigCodeBench Hard O	Code generation (hard)	Harder variant of BigCodeBench testing complex programming and library tasks with function-level code generation.	★★★★★		🇺🇸 Claude 3.7 Sonnet (2025-02-19)	35.8%
46	BLINK O	Multimodal grounding	Evaluates visual-language grounding and reference resolution to reduce hallucinations.	★★★★★		🇨🇳 Seed1.5-VL-Thinking	72.4%
47	BoB-HVR O	Composite capability index	Hard, Versatile, and Relevant composite score across eight capability buckets.	★★★★★		🇺🇸 Llama 3 70B	9.0%
48	BOLD o	Bias evaluation	Bias in Open-ended Language Dataset probing demographic biases in text generation.	★★★★★		🇫🇷 Mixtral 8x 7B	↓ 0.1%
49	BoolQ O	Reading comprehension	Yes/no QA from naturally occurring questions.	★★★★★	171	🇺🇸 Gemma 2 27B	84.8%
50	BrowseComp O	Web browsing	Web browsing comprehension and competence benchmark.	★★★★★		🇺🇸 GPT-5	54.9%
51	BrowseComp_zh O	Web browsing (Chinese)	Chinese variant of the BrowseComp web browsing benchmark.	★★★★★		🇺🇸 o3	58.1%
52	BRuMo25 o	Math competition	BruMo 2025 olympiad-style mathematics benchmark.	★★★★★		🇺🇸 QuestA Nemotron 1.5B	69.5%
53	BuzzBench O	Humor analysis	A humour analysis benchmark.	★★★★★		🇺🇸 Gemini 2.5 Pro	71.1%
54	C-Eval O	Chinese exams	Comprehensive Chinese exam benchmark across multiple subjects.	★★★★★		🇨🇳 Kimi-K2 Base	92.5%
55	C3-Bench o	Reasoning (Chinese)	Comprehensive Chinese reasoning capability benchmark.	★★★★★	35	🇨🇳 GLM-4.5 Base	83.1%
56	CaseLaw v2 O	Legal reasoning	U.S. case law benchmark evaluating legal reasoning and judgment over court opinions.	★★★★★		🇺🇸 GPT-4.1	78.1%
57	CC-OCR o	OCR (cross-lingual)	Cross-lingual OCR benchmark evaluating character recognition across mixed-language documents.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	81.5%
58	CFEval o	Coding ELO / contest eval	Contest-style coding evaluation with ELO-like scoring.	★★★★★		🇨🇳 Qwen3-235B-A22B-Thinking-2507	2134
59	Charades-STA O	Video grounding	Charades-STA temporal grounding (mIoU).	★★★★★		🇨🇳 Seed1.5-VL-Thinking	64.0%
60	ChartMuseum o	Chart understanding	Large-scale curated collection of charts for evaluating parsing, grounding, and reasoning.	★★★★★		🇺🇸 GPT-5 mini	63.3%
61	ChartQA O	Chart understanding (VQA)	Visual question answering over charts and plots.	★★★★★		G MiMo-VL 7B-SFT	92.9%
62	ChartQA-Pro o	Chart understanding (VQA)	Professional-grade chart question answering with diverse chart types and complex reasoning.	★★★★★		🇨🇳 GLM-4.5V	64.0%
63	CharXiv (DQ) O	Chart description (PDF)	Scientific chart/table descriptive questions from arXiv PDFs.	★★★★★		🇺🇸 o3-high	95.0%
64	CharXiv (RQ) O	Chart reasoning (PDF)	Scientific chart/table reasoning questions from arXiv PDFs.	★★★★★		🇺🇸 GPT-5	81.1%
65	Chinese SimpleQA o	QA (Chinese)	Chinese variant of the SimpleQA benchmark.	★★★★★		🇨🇳 Kimi-K2 Base	77.6%
66	CLUEWSC o	Coreference reasoning (Chinese)	Chinese Winograd Schema-style coreference benchmark from CLUE.	★★★★★		🇨🇳 DeepSeek R1	92.8%
67	CMath o	Math (Chinese)	Chinese mathematics benchmark.	★★★★★		🇨🇳 ERNIE 4.5 424B A47B	96.7%
68	CMMLU o	Chinese multi-domain	Chinese counterpart to MMLU.	★★★★★	781	🇨🇳 Qwen2.5 Max	91.9%
69	Codeforces O	Competitive programming	Competitive programming performance on Codeforces problems (ELO).	★★★★★		🇺🇸 o4 mini	2719
70	COLLIE o	Instruction following	Comprehensive instruction-following evaluation suite.	★★★★★	55	🇺🇸 GPT-5	99.0%
71	CommonsenseQA O	Commonsense QA	Multiple-choice QA requiring commonsense knowledge.	★★★★★		🇺🇸 Llama 3.1 405B Base	85.8%
72	CountBench O	Visual counting	Object counting and numeracy benchmark for visual-language models across varied scenes.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	93.7%
73	CountBenchQA O	Visual counting QA	Visual question answering benchmark focused on counting objects across varied scenes.	★★★★★		G Moondream-9B-A2B	93.2%
74	CRAG o	Retrieval QA	Complex Retrieval-Augmented Generation benchmark for grounded question answering.	★★★★★		G Jamba Mini 1.6	76.2%
75	Creative Story‑Writing Benchmark V3 O	Creative writing	Story writing benchmark evaluating creativity, coherence, and style (v3).	★★★★★	291	🇨🇳 Kimi-K2-Instruct-0905	8.7%
76	Longform Creative Writing O	Creative writing	Longform creative writing evaluation (EQ-Bench).	★★★★★	20	🇨🇳 DeepSeek V3 0528	78.9%
77	Creative Writing v3 O	Creative writing	A LLM-judged creative writing benchmark.	★★★★★	54	🇺🇸 o3	1661
78	CRUX-I O	Code reasoning	Code Reasoning and Understanding eXam – Interactive.	★★★★★		🇺🇸 GPT-4 Turbo (2024-04-09) CoT	75.7%
79	CRUX-O O	Code reasoning	Code Reasoning and Understanding eXam – Offline.	★★★★★		🇺🇸 GPT-4 0613 CoT	88.2%
80	CruxEval O	Code reasoning	Mathematical coding challenge set from the CruxEval benchmark.	★★★★★		🇨🇳 Qwen3-32B	78.5%
81	CV-Bench O	Computer vision QA	Diverse CV tasks for VLMs.	★★★★★		🇨🇳 Seed1.5-VL-Thinking	89.7%
82	DeepMind Mathematics o	Math reasoning	Synthetic math problem sets from DeepMind covering arithmetic, algebra, calculus, and more.	★★★★★		G Granite-4.0-H-Small	59.3%
83	Design2Code O	Coding (UI)	Translating UI designs into code.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	93.4%
84	DesignArena O	Generative design	Leaderboard tracking generative design systems across layout, branding, and marketing tasks.	★★★★★		🇺🇸 Claude Sonnet 4.5 (Thinking)	1410
85	DetailBench o	Spot small mistakes	Evaluates whether LLMs can notice subtle errors and minor inconsistencies in text.	★★★★★		🇺🇸 Llama 4 Maverick	8.7%
86	DocVQA O	Document understanding (VQA)	Visual question answering over scanned documents.	★★★★★		🇨🇳 Seed1.5-VL-Thinking	96.9%
87	DROP O	Reading + reasoning	Discrete reasoning over paragraphs (addition, counting, comparisons).	★★★★★		🇨🇳 DeepSeek R1	92.2%
88	DynaMath O	Math reasoning (video)	Dynamic/video-based mathematical reasoning evaluating temporal and visual understanding.	★★★★★		🇺🇸 GPT-4o	63.7%
89	Economically important tasks o	Industry QA (cross-domain)	Evaluation suite of real-world, economically impactful tasks across key industries and workflows.	★★★★★		🇺🇸 GPT-5	47.1%
90	EgoSchema O	Egocentric video QA	EgoSchema validation accuracy.	★★★★★		🇨🇳 Qwen2-VL 72B Instruct	77.9%
91	EmbSpatialBench o	Spatial understanding	Embodied spatial understanding benchmark evaluating navigation and localization.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	84.3%
92	Enterprise RAG o	Retrieval-augmented generation	Enterprise retrieval-augmented generation evaluation covering internal knowledge bases.	★★★★★		🇺🇸 Apriel Nemotron 15B Thinker	69.2%
93	EQ-Bench O	Reasoning	General reasoning benchmark assessing equation/logic capabilities.	★★★★★	352	G Jan v1 2509	85.0%
94	EQ-Bench 3 O	Emotional intelligence (roleplay)	A benchmark measuring emotional intelligence in challenging roleplays, judged by Sonnet 3.7.	★★★★★	21	🇨🇳 Kimi K2 Instruct	1555
95	ERQA O	Spatial reasoning	Spatial recognition and reasoning QA benchmark (ERQA).	★★★★★		🇺🇸 GPT-5	65.7%
96	EvalPerf O	Code evaluation performance	Measures performance of LLM code evaluation, including runtime, memory, and efficiency metrics.	★★★★★		🇺🇸 GPT-4o (2024-08-06)	100.0%
97	EvalPlus O	Code generation	Aggregated code evaluation suite from EvalPlus.	★★★★★	1577	🇺🇸 o1 Mini	89.0%
98	FACTS Grounding o	Grounding / factuality	Grounded factuality benchmark evaluating model alignment with source facts.	★★★★★		🇺🇸 Gemini 2.5 Pro	87.8%
99	FActScore o	Hallucination rate on open-source prompts	Measures hallucination rate on an open-source prompt suite; lower is better.	★★★★★		🇺🇸 GPT-5	↓ 1.0%
100	FAIX Agent O	Composite capability index		★★★★★		🇺🇸 Holo1.5-72B	90.7%
101	FAIX Code O	Composite capability index		★★★★★		🇺🇸 Claude Sonnet 4.5	74.4%
102	FAIX Math O	Composite capability index		★★★★★		🇨🇳 Qwen3-VL-235B-A22B Instruct	100.0%
103	FAIX OCR O	Composite capability index		★★★★★		🇺🇸 o3 (Low)	86.0%
104	FAIX Safety O	Composite safety index		★★★★★		🇺🇸 Claude Sonnet 4.5	74.0%
105	FAIX STEM O	Composite capability index		★★★★★		🇨🇳 Qwen3-VL-235B-A22B Instruct	100.0%
106	FAIX Text O	Composite capability index		★★★★★		🇨🇳 GLM-4.5V	70.0%
107	FAIX Visual O	Composite capability index		★★★★★		🇺🇸 Holo1.5-72B	91.6%
108	FAIX Writing O	Composite capability index		★★★★★		🇨🇳 Qwen3 235B A22B Instruct 2507	69.3%
109	FinanceReasoning o	Financial reasoning	Financial reasoning benchmark evaluating quantitative and qualitative finance problem solving.	★★★★★		G Ling 1T	87.5%
110	FinanceAgent o	Agentic finance tasks	Interactive financial agent benchmark requiring multi-step tool use.	★★★★★		🇺🇸 Claude Sonnet 4.5	55.3%
111	FinanceBench (FullDoc) o	Finance QA	FinanceBench full-document question answering benchmark requiring long-context financial understanding.	★★★★★		G Jamba Mini 1.6	45.4%
112	FinSearchComp O	Financial retrieval	Financial search and comprehension benchmark measuring retrieval grounded reasoning over financial content.	★★★★★		🇺🇸 Grok 4	68.9%
113	FinSearchComp-CN O	Financial retrieval (Chinese)	Chinese financial search and comprehension benchmark measuring retrieval-grounded reasoning over regional financial content.	★★★★★		G doubao-1-5-vision-pro	54.2%
114	Flame-React-Eval o	Frontend coding	Front-end React coding tasks and evaluation.	★★★★★		🇨🇳 GLM-4.5V	82.5%
115	FRAMES o	Interactive reasoning	Frame-based interactive reasoning and dialogue benchmark.	★★★★★		🇨🇳 Tongyi DeepResearch	90.6%
116	FreshQA o	Recency QA	Question answering benchmark emphasizing up-to-date knowledge and recency.	★★★★★		🇨🇳 Qwen3-4B Thinking 2507	66.9%
117	FullStackBench O	Full-stack development	End-to-end web/app development tasks and evaluation.	★★★★★		🇺🇸 GPT-4.1	68.5%
118	GAIA o	General AI tasks	Comprehensive benchmark for agentic tasks.	★★★★★		🇨🇳 Tongyi DeepResearch	70.9%
119	GAIA 2 O	General agent tasks	Grounded agentic intelligence benchmark version 2 covering multi-tool tasks.	★★★★★		🇺🇸 GPT-5 High	42.1%
120	GDPVal o	General capability	GDPVal benchmark evaluating broad general capabilities of LLMs across diverse tasks.	★★★★★		🇺🇸 Claude Opus 4.1	47.6%
121	GeoBench1 o	Geospatial reasoning	Geospatial visual QA and reasoning (set 1).	★★★★★		🇨🇳 GLM-4.5V	79.7%
122	Global-MMLU o	Multi-domain knowledge (global)	Full Global-MMLU evaluation across diverse languages and regions.	★★★★★		🇺🇸 Llama 3.3 70B	77.8%
123	Gorilla Benchmark API Bench o	Tool use	Gorilla API Bench tool-use evaluation.	★★★★★		🇺🇸 Llama 3.1 405B	35.3%
124	GPQA O	Graduate-level QA	Graduate-level question answering evaluating advanced reasoning.	★★★★★	406	🇺🇸 Grok 4	88.4%
125	GPQA-diamond O	Graduate-level QA	Hard subset of GPQA (diamond level).	★★★★★		🇺🇸 GPT-5 pro	89.4%
126	Ground-UI-1K O	GUI grounding	Accuracy on the Ground-UI-1K grounding benchmark.	★★★★★		🇨🇳 Qwen2.5-VL 72B	85.4%
127	GSM-Plus O	Math (grade-school, enhanced)	Enhanced GSM-style grade-school math benchmark variant.	★★★★★		🇨🇳 Qwen3-4B	82.1%
128	GSM-Symbolic o	Math reasoning	Symbolic reasoning variant of GSM that tests algebraic manipulation and arithmetic with structured problems.	★★★★★		G Granite-4.0-H-Small	87.4%
129	GSM8K O	Math (grade-school)	Grade-school math word problems requiring multi-step reasoning.	★★★★★	1322	🇨🇳 Kimi K2 Instruct	97.3%
130	GSM8K (DE) o	Math (grade-school, German)	German translation of the GSM8K grade-school math word problems.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	0.6%
131	GSO Benchmark O	Code generation	LiveCodeBench GSO benchmark.	★★★★★		🇺🇸 o3-high	8.8%
132	HallusionBench O	Multimodal hallucination	Benchmark for evaluating hallucination tendencies in multimodal LLMs.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	66.7%
133	HarmfulQA o	Safety	Harmful question set testing models' ability to avoid unsafe answers.	★★★★★	104	G K2-THINK	99.0%
134	HealthBench O	Medical QA	Comprehensive medical knowledge and clinical reasoning benchmark across specialties and tasks.	★★★★★		🇺🇸 GPT-5	67.2%
135	HealthBench-Hard o	Medical QA (hard)	Challenging subset of HealthBench focusing on complex, ambiguous clinical cases.	★★★★★		🇺🇸 GPT-5	46.2%
136	HealthBench-Hard Hallucinations o	Medical hallucination safety	Measures hallucination and unsafe medical advice under hard clinical scenarios.	★★★★★		🇺🇸 GPT-5	↓ 1.6%
137	HellaSwag O	Commonsense reasoning	Adversarial commonsense sentence completion.	★★★★★	220	🇨🇳 DeepSeek V3 Base	96.4%
138	HellaSwag (DE) o	Commonsense reasoning (German)	German translation of the HellaSwag commonsense benchmark.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	0.7%
139	HELMET LongQA o	Long-context QA	Long-context subset of the HELMET benchmark focusing on grounded question answering.	★★★★★		G Jamba Mini 1.6	46.9%
140	HeroBench O	Long-horizon planning	Benchmark for Long-Horizon Planning and Structured Reasoning in Virtual Worlds.	★★★★★		🇺🇸 Grok 4	91.7%
141	HHEM v2.1 O	Hallucination detection	Hughes Hallucination Evaluation Model (Vectara) — lower is better.	★★★★★		G AntGroup Finix_S1_32b	↓ 0.6%
142	HLE O	Multi-domain reasoning	Challenging LLMs at the frontier of human knowledge.	★★★★★	1085	🇨🇳 Tongyi DeepResearch	32.9%
143	HMMT o	Math (competition)	Harvard–MIT Mathematics Tournament problems.	★★★★★		🇺🇸 GPT-5 pro	100.0%
144	HMMT 2025 O	Math (competition)	Harvard–MIT Mathematics Tournament 2025 problems.	★★★★★		🇺🇸 Grok 4 Fast	93.3%
145	HMMT25 o	Math (competition)	Harvard-MIT Mathematics Tournament 2025 benchmark.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Thinking	67.6%
146	HRBench 4K o	Hallucination robustness	Hallucination robustness benchmark with 4K token contexts.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Instruct	89.5%
147	HRBench 8K o	Hallucination robustness	Hallucination robustness benchmark with 8K token contexts.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Instruct	82.5%
148	HumanEval O	Code generation	Python synthesis problems evaluated by unit tests.	★★★★★	2916	🇺🇸 o1-preview	96.3%
149	HumanEval+ O	Code generation	Extended HumanEval with more tests.	★★★★★	1577	🇺🇸 Claude Sonnet 4	94.5%
150	Hypersim o	3D scene understanding	Hypersim benchmark for synthetic indoor scene understanding and reconstruction.	★★★★★		🇺🇸 GPT-5 Mini Minimal	39.3%
151	IFBench O	Instruction following	Instruction-following benchmark measuring compliance and adherence.	★★★★★	70	🇫🇷 Mistral Small 3.2 24B Instruct	84.8%
152	IFEval O	Instruction following	Instruction following capability evaluation for LLMs.	★★★★★	36312	🇺🇸 o3 mini-high	93.9%
153	INCLUDE O	Inclusiveness / bias	Evaluates inclusive language use and bias mitigation in model outputs.	★★★★★		🇺🇸 Gemini-2.5-Flash Thinking	83.9%
154	InfoQA O	Information-seeking QA	Information retrieval question answering benchmark evaluating factual responses.	★★★★★		🇨🇳 Qwen2-VL-72B	84.5%
155	InfoVQA O	Infographic VQA	Visual question answering over infographics requiring reading, counting, and reasoning.	★★★★★		🇨🇳 Seed1.5-VL-Thinking	91.2%
156	JudgeMark v2.1 O	LLM judging ability	A benchmark measuring LLM judging ability.	★★★★★		🇺🇸 Claude Sonnet 4	82.0%
157	KMMLU-Pro O	Multilingual knowledge	Korean Multilingual Massive Multitask Language Understanding Pro	★★★★★		🇺🇸 o1	77.5%
158	KMMLU-Redux O	Multilingual knowledge	Redux variant of KMMLU benchmark	★★★★★		🇺🇸 o1	81.1%
159	KOR-Bench o	Reasoning	Comprehensive reasoning benchmark spanning diverse domains and cognitive skills.	★★★★★		G Ling 1T	76.0%
160	KSM o	Multilingual math	Korean STEM and math benchmark	★★★★★		G EXAONE Deep 2.4B	60.9%
161	LAMBADA o	Language modeling	Word prediction requiring broad context understanding.	★★★★★		🇺🇸 GPT-3	86.4%
162	LatentJailbreak o	Safety / jailbreak	Robustness to latent jailbreak adversarial techniques.	★★★★★	39	🇺🇸 GPT-3.5-turbo	77.4%
163	LiveBench O	General capability	Continually updated capability benchmark across diverse tasks.	★★★★★		🇺🇸 Gemini 2.5 Pro	82.4%
164	LiveBench 20241125 O	General capability	LiveBench snapshot (2024-11-25) tracking mixed-task evals.	★★★★★		🇺🇸 Gemini 2.5 Pro	82.4%
165	LiveCodeBench O	Code generation	Live coding and execution-based evaluation benchmark (v6 dataset).	★★★★★		🇺🇸 GPT-5 mini	86.6%
166	LiveCodeBench v5 (2024.10-2025.02) O	Code generation	LiveCodeBench v5 snapshot covering Oct 2024-Feb 2025.	★★★★★		🇨🇳 Qwen3-235B-A22B	70.7%
167	LiveMCP-101 O	Agent real-time eval	A novel real-time evaluation framework and benchmark to stress‑test agents on complex, real‑world tasks.	★★★★★		🇺🇸 GPT-5	58.4%
168	LMArena Text O	Crowd eval (text)	Chatbot Arena text evaluation (average win rate).	★★★★★		🇺🇸 Gemini 2.5 Pro	1455
169	LMArena Vision O	Crowd eval (vision)	Chatbot Arena vision evaluation leaderboard (ELO ratings).	★★★★★		🇺🇸 Gemini 2.5 Pro	1242
170	LogicVista O	Visual logical reasoning	Visual logic and pattern reasoning tasks requiring compositional and spatial understanding.	★★★★★		🇨🇳 GLM-4.5V	62.4%
171	LogiQA o	Logical reasoning	Reading comprehension with logical reasoning.	★★★★★	138	G Pythia 70M	23.5%
172	LongBench o	Long-context eval	Long-context understanding across tasks.	★★★★★	957	G Jamba Mini 1.6	32.0%
173	LongFact-Concepts o	Hallucination rate on open-source prompts	Long-context factuality eval focused on conceptual statements; lower is better.	★★★★★		🇺🇸 GPT-5	↓ 0.7%
174	LongFact-Objects o	Hallucination rate on open-source prompts	Long-context factuality eval focused on object/entity references; lower is better.	★★★★★		🇺🇸 GPT-5	↓ 0.8%
175	LVBench O	Video understanding	Long video understanding benchmark (LVBench).	★★★★★		🇺🇸 Gemini 2.5 Pro	73.0%
176	M3GIA (CN) o	Chinese multimodal QA	Chinese-language M3GIA benchmark covering grounded multimodal question answering.	★★★★★		🇨🇳 Seed1.5-VL-Thinking	91.2%
177	Mantis O	Multimodal reasoning	Multimodal reasoning and instruction following benchmark (Mantis).	★★★★★		G dots.vlm1	86.2%
178	MASK O	Safety / red teaming	Model behavior safety assessment via red-teaming scenarios.	★★★★★		🇺🇸 Claude Sonnet 4 (t)	95.3%
179	MATH O	Math (competition)	Competition-level mathematics across algebra, geometry, number theory, combinatorics.	★★★★★	1185	🇺🇸 o3 mini	97.9%
180	MATH Level 5 o	Math (competition)	Level 5 subset of the MATH benchmark emphasizing the hardest competition-style problems.	★★★★★		🇨🇳 Qwen3-4B-Instruct-2507	73.6%
181	MATH500 O	Math (competition)	500 curated math problems for evaluating high-level reasoning.	★★★★★		🇺🇸 GPT-5	99.2%
182	MATH500 (ES) o	Math (multilingual)	Spanish MATH500 benchmark	★★★★★		G EXAONE 4.0 1.2B	88.8%
183	MathVerse-mini o	Math reasoning (multimodal)	Compact MathVerse split focusing on single-image math puzzles and visual reasoning.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	85.0%
184	MathVerse-Vision O	Math reasoning (multimodal)	Multi-image visual mathematical reasoning tasks from the MathVerse ecosystem.	★★★★★		🇨🇳 GLM-4.5V	72.1%
185	MathVision O	Math reasoning (multimodal)	Visual math reasoning benchmark with problems that combine images (charts, diagrams) and text.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	74.6%
186	MathVista O	Multimodal math reasoning	Visual math reasoning across diverse tasks.	★★★★★		🇺🇸 o3	86.8%
187	MathVista-Mini O	Math reasoning (multimodal)	Lightweight subset of MathVista for quick evaluation of visual mathematical reasoning.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	85.8%
188	MBPP O	Code generation	Short Python problems with hidden tests.	★★★★★	36312	🇺🇸 o1-preview	95.5%
189	MBPP+ O	Code generation	Extended MBPP with more tests and stricter evaluation.	★★★★★		🇺🇸 Llama 3.1 405B	88.6%
190	MCP Universe O	Agent evaluation	Benchmarks multi-step tool-use agents across diverse task suites with a unified overall success metric.	★★★★★		🇺🇸 GPT-5 High	44.2%
191	MCPMark O	Agent tool-use (MCP)	Benchmark for Model Context Protocol (MCP) agent tool-use.	★★★★★	127	🇺🇸 GPT-5	46.9%
192	MGSM O	Math (multilingual)	Multilingual grade school math word problems.	★★★★★		🇺🇸 Claude Opus 4.1 (2025-08-05) Thinking	94.4%
193	MIABench o	Multimodal instruction following	Multimodal instruction-following benchmark evaluating accuracy on complex image-text tasks.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	92.7%
194	Minerva Math o	University-level math	Advanced quantitative reasoning set inspired by the Minerva benchmark for STEM problem solving.	★★★★★		G Granite-4.0-H-Small	74.0%
195	MiniF2F (Test) o	Math competition	MiniF2F competition benchmark (test split).	★★★★★		🇨🇳 LongCat-Flash-Thinking	81.6%
196	MixEval o	Multi-task reasoning	Mixed-subject benchmark covering knowledge and reasoning tasks across domains.	★★★★★		🇺🇸 o1 Mini	82.9%
197	MixEval Hard o	Multi-task reasoning (hard)	Hard subset of MixEval covering diverse reasoning tasks.	★★★★★		🇨🇳 Qwen3-4B	31.6%
198	MLVU O	Large video understanding	MLVU: Large-scale multi-task benchmark for video understanding.	★★★★★		🇺🇸 GPT-5	86.2%
199	MM-MT-Bench O	Multimodal instruction following	Multi-turn multimodal instruction following benchmark evaluating dialogue quality and helpfulness.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	8.5%
200	MMBench v1.1 (EN dev) O	General VQA	English dev split of MMBench v1.1 measuring multimodal question answering.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	90.6%
201	MMBench v1.1 (CN) O	Multimodal understanding (Chinese)	MMBench v1.1 Chinese subset for evaluating multimodal LLMs.	★★★★★		🇨🇳 Keye-VL 8B	89.8%
202	MMBench v1.1 (EN) O	Multimodal understanding (English)	MMBench v1.1 English subset for evaluating multimodal LLMs.	★★★★★		🇨🇳 Keye-VL 8B	89.7%
203	MME-RealWorld (cn) o	Real-world perception (CN)	MME-RealWorld Chinese split.	★★★★★		🇺🇸 GPT-4o	58.5%
204	MME-RealWorld (en) o	Real-world perception (EN)	MME-RealWorld English split.	★★★★★		G MiMo-VL 7B-RL	59.1%
205	MMLongBench-Doc O	Long-context multimodal documents	Evaluates long-context document understanding with mixed text, tables, and figures across multiple pages.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	56.2%
206	MMLU O	Multi-domain knowledge	57 tasks spanning STEM, humanities, social sciences; broad knowledge and reasoning.	★★★★★	1488	🇺🇸 GPT-5	93.5%
207	MMLU (cloze) o	Multi-domain knowledge (cloze)	Cloze-form MMLU evaluation variant.	★★★★★		🇺🇸 SmolLM2 135M Base	31.5%
208	Full Text MMLU o	Multi-domain knowledge (long-form)	Full-context MMLU variant evaluating reasoning over long passages.	★★★★★		🇺🇸 Llama 3.3 70B Instruct	0.8%
209	MMLU-Pro O	Multi-domain knowledge	Harder successor to MMLU with more challenging questions.	★★★★★	286	🇺🇸 o1	89.3%
210	MMLU Pro MCF o	Multi-domain knowledge (few-shot)	MMLU-Pro common format (MCF) few-shot evaluation.	★★★★★		🇨🇳 Qwen3-4B-Base	41.1%
211	MMLU-ProX O	Multi-domain knowledge	Cross-lingual and robust variant of MMLU-Pro.	★★★★★		🇨🇳 Qwen3-235B-A22B-Thinking-2507	81.0%
212	MMLU-Redux O	Multi-domain knowledge	Updated MMLU-style evaluation with revised questions and scoring.	★★★★★		🇨🇳 Qwen3-235B-A22B-Thinking-2507	93.8%
213	MMLU-STEM O	STEM knowledge	STEM subset of MMLU.	★★★★★	1488	G Falcon-H1-34B-Instruct	83.6%
214	MMMLU O	Multi-domain knowledge (multilingual)	Massively multilingual MMLU-style evaluation across many languages.	★★★★★		🇺🇸 Claude Opus 4.1	89.5%
215	MMMLU (ES) o	Multilingual knowledge	Spanish MMMLU benchmark	★★★★★		🇺🇸 SmolLM 3 3B	64.7%
216	MMMU O	Multimodal understanding	Multi-discipline multimodal understanding benchmark.	★★★★★		🇺🇸 Gemini 2.5 Pro	84.2%
217	MMMU PRO O	Multimodal understanding (hard)	Professional/advanced subset of MMMU for multimodal reasoning.	★★★★★		🇺🇸 GPT-5	78.4%
218	MMMU-Pro (vision) o	Multimodal understanding (vision)	MMMU-Pro vision-only setting.	★★★★★		🇺🇸 Claude 3.7 Sonnet	45.8%
219	MMStar O	Multimodal reasoning	Broad evaluation of multimodal LLMs across diverse tasks.	★★★★★		🇺🇸 Gemini 2.5 Pro	78.7%
220	MMVP o	Multimodal video perception	Benchmark for multimodal video understanding and perception.	★★★★★		🇨🇳 R-4B-RL	80.7%
221	MMVU o	Video understanding	Multimodal video understanding benchmark (MMVU).	★★★★★		🇨🇳 GLM-4.5V	68.7%
222	MotionBench o	Video motion understanding	Video motion and temporal reasoning benchmark.	★★★★★		🇨🇳 GLM-4.5V	62.4%
223	MT-Bench O	Chat ability	Multi-turn chat evaluation via GPT-4 grading.	★★★★★	39074	🇺🇸 Apriel Nemotron 15B Thinker	85.7%
224	MTOB (full book) o	Long-form reasoning	Long-context book understanding benchmark (full-book setting).	★★★★★		🇺🇸 Llama 4 Maverick	50.8%
225	MTOB (half book) o	Long-form reasoning	Long-context book understanding benchmark (half-book setting).	★★★★★		🇺🇸 Llama 4 Maverick	54.0%
226	MUIRBENCH O	Multimodal robustness	Evaluates multimodal understanding robustness and reliability.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	80.1%
227	Multi-IF O	Instruction following (multi-task)	Composite instruction-following evaluation across multiple tasks.	★★★★★		🇺🇸 o3 mini-high	79.5%
228	Multi-IFEval O	Instruction following (multi-task)	Multi-task variant of instruction-following evaluation.	★★★★★		🇺🇸 Llama 3.3 70B	88.7%
229	Multi-SWE-Bench o	Code repair (multi-repo)	Multi-repository SWE-Bench variant.	★★★★★	246	🇺🇸 Claude Sonnet 4	35.7%
230	MultiChallenge o	Multi-task reasoning	Composite benchmark across diverse challenges by Scale AI.	★★★★★		🇺🇸 GPT-5	69.6%
231	MultiPL-E O	Code generation (multilingual)	Multilingual code generation and execution benchmark across many programming languages.	★★★★★	269	🇨🇳 Qwen3-235B-A22B	87.9%
232	MultiPL-E HumanEval o	Code generation (multilingual)	MultiPL-E variant of HumanEval tasks.	★★★★★		🇺🇸 Llama 3.1 405B	75.2%
233	MultiPL-E MBPP o	Code generation (multilingual)	MultiPL-E variant of MBPP tasks.	★★★★★		🇺🇸 Llama 3.1 405B	65.7%
234	MuSR O	Reasoning	Multistep Soft Reasoning.	★★★★★		🇨🇳 ERNIE 4.5 424B A47B	69.9%
235	MVBench O	Video QA	Multi-view or multi-video QA benchmark (MVBench).	★★★★★		🇨🇳 GLM-4.5V	73.0%
236	Natural2Code o	Code generation	Natural language to code benchmark for instruction-following synthesis.	★★★★★		🇺🇸 Gemini 2.0 Flash	92.9%
237	NaturalQuestions O	Open-domain QA	Google NQ; real user questions with long/short answers.	★★★★★		🇫🇷 Mixtral 8x22B	40.1%
238	Nexus (0-shot) o	Tool use	Nexus tool-use benchmark, zero-shot setting.	★★★★★		🇺🇸 Llama 3.1 405B	58.7%
239	Objectron o	Object detection	Objectron benchmark for 3D object detection in video captures.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	71.2%
240	OCRBench O	OCR (vision text extraction)	Optical character recognition benchmark evaluating text extraction from images, documents, and complex layouts.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Instruct	90.3%
241	OCRBenchV2 (CN) o	OCR (Chinese)	OCRBenchV2 Chinese subset assessing OCR performance on Chinese-language documents.	★★★★★		🇨🇳 Qwen2.5-VL 72B Instruct	63.7%
242	OCRBenchV2 (EN) o	OCR (English)	OCRBenchV2 English subset evaluating OCR accuracy on English documents and layouts.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	66.8%
243	OCRReasoning o	OCR reasoning	OCR reasoning benchmark combining text extraction with multi-step reasoning over documents.	★★★★★		🇺🇸 Gemini 2.5 Pro	70.8%
244	ODinW-13 o	Object detection (in the wild)	Object Detection in the Wild benchmark covering 13 real-world domains.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Instruct	47.5%
245	OJBench O	Code generation (online judge)	Programming problems evaluated via online judge-style execution.	★★★★★		🇺🇸 Gemini 2.5 Pro	41.6%
246	OlympiadBench o	Math (olympiad)	Advanced mathematics olympiad-style problem benchmark.	★★★★★		🇨🇳 Hunyuan-7B-Instruct	76.5%
247	OlympicArena o	Math (competition)	Olympiad-style mathematics reasoning benchmark.	★★★★★		🇨🇳 DeepSeek V3	76.2%
248	Omni-MATH O	Math reasoning	Omni-MATH benchmark covering diverse math reasoning tasks across difficulty levels.	★★★★★		G Ling 1T	74.5%
249	Omni-MATH-HARD O	Math	Challenging math benchmark (Omni-MATH-HARD).	★★★★★		🇺🇸 GPT-5 High	73.6%
250	OmniSpatial o	Spatial reasoning	Spatial understanding and reasoning benchmark (OmniSpatial).	★★★★★		🇨🇳 GLM-4.5V	51.0%
251	OpenBookQA O	Science QA	Open-book multiple choice science questions with supporting facts.	★★★★★	128	🇨🇳 Qwen3	96.4%
252	OpenRewrite-Eval o	Rewrite quality	OpenRewrite evaluation; micro-averaged RougeL.	★★★★★		🇨🇳 Qwen2.5 1.5B Instruct	46.9%
253	OptMATH o	Math optimization reasoning	OptMATH benchmark targeting challenging math optimization and problem-solving tasks.	★★★★★		G Ling 1T	57.7%
254	OSWorld o	GUI agents	Agentic GUI task completion and grounding on desktop environments.	★★★★★		🇺🇸 Claude Sonnet 4.5	61.4%
255	OSWorld-G O	GUI agents	OSWorld-G center accuracy (no_refusal).	★★★★★		🇺🇸 Holo1.5-72B	71.8%
256	OSWorld2 o	GUI agents	Second-generation OSWorld GUI agent benchmark.	★★★★★		🇨🇳 GLM-4.5V	35.8%
257	PIQA O	Physical commonsense	Physical commonsense about everyday tasks and object affordances.	★★★★★		🇨🇳 GLM-4.5 Base	87.1%
258	PixmoCount O	Visual counting	Counting objects/instances in images (PixmoCount).	★★★★★		🇺🇸 Molmo-72B	85.2%
259	PolyMATH O	Math reasoning	Polyglot mathematics benchmark assessing cross-topic math reasoning.	★★★★★		🇨🇳 Qwen3-235B-A22B-Thinking-2507	60.1%
260	POPE o	Hallucination detection	Vision-language hallucination benchmark focusing on object existence verification.	★★★★★		G Moondream-9B-A2B	89.0%
261	PopQA O	Knowledge / QA	Open-domain popular culture question answering benchmark testing long-tail factual recall.	★★★★★		🇺🇸 Llama 3.1 8B Instruct	28.8%
262	QuAC o	Conversational QA	Question answering in context.	★★★★★		🇺🇸 Llama 3.1 405B Base	53.6%
263	QuALITY o	Long-context reading comprehension	Long-document multiple-choice reading comprehension benchmark.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	0.5%
264	RACE o	Reading comprehension	English exams for middle and high school.	★★★★★		G RND1-Base-0910	57.6%
265	RealWorldQA O	Real-world visual QA	Visual question answering with real-world images and scenarios.	★★★★★		🇺🇸 GPT-5	82.8%
266	RefCOCO O	Referring expressions	RefCOCO average accuracy at IoU 0.5 (val).	★★★★★		🇨🇳 InternVL3.5-4B	92.4%
267	RefCOCOg o	Referring expressions	RefCOCOg average accuracy at IoU 0.5 (val).	★★★★★		G Moondream-9B-A2B	88.6%
268	RefCOCO+ o	Referring expressions	RefCOCO+ accuracy at IoU 0.5 on the val split.	★★★★★		G Moondream-9B-A2B	81.8%
269	RefSpatialBench o	Spatial reasoning	Reference spatial understanding benchmark covering spatial grounding tasks.	★★★★★		🇨🇳 Qwen2.5-VL 72B Instruct	72.1%
270	RepoBench O	Code understanding	Repository-level code comprehension and reasoning benchmark.	★★★★★		🇺🇸 Claude Sonnet 4.5	83.8%
271	RoboSpatialHome o	Embodied spatial understanding	RoboSpatialHome benchmark for embodied spatial reasoning in domestic environments.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	73.9%
272	Roo Code Evals O	Code assistant eval	Community-maintained coding evals and leaderboard by Roo Code.	★★★★★		🇺🇸 GPT-5 mini	99.0%
273	Ruler 128k o	Long-context eval	RULER benchmark at 128k context window.	★★★★★		🇫🇷 Mistral Medium 3	90.2%
274	Ruler 32k o	Long-context eval	RULER benchmark at 32k context window.	★★★★★		🇫🇷 Mistral Medium 3	96.0%
275	SALAD-Bench o	Safety alignment	Safety Alignment and Dangerous-behavior benchmark evaluating harmful assistance and refusal consistency.	★★★★★		G Granite-4.0-H-Micro	↓ 96.8%
276	SciCode (sub) O	Code	SciCode subset score (sub).	★★★★★		🇺🇸 Grok 4	45.7%
277	SciCode (main) O	Code	SciCode main score.	★★★★★		🇺🇸 Gemini 2.5 Pro	15.4%
278	ScienceQA O	Science QA (multimodal)	Multiple-choice science questions with images, diagrams, and text context.	★★★★★		G FastVLM-7B	96.7%
279	SciQ o	Science QA	Multiple choice science questions.	★★★★★		G Pythia 12B	92.9%
280	ScreenQA Complex O	GUI QA	Complex ScreenQA benchmark accuracy.	★★★★★		🇺🇸 Holo1.5-72B	87.1%
281	ScreenQA Short O	GUI QA	Short-form ScreenQA benchmark accuracy.	★★★★★		🇺🇸 Holo1.5-72B	91.9%
282	ScreenSpot O	Screen UI locators	Center accuracy on ScreenSpot.	★★★★★		🇨🇳 Qwen3-VL-235B-A22B Thinking	95.4%
283	ScreenSpot-Pro O	Screen UI locators	Average center accuracy on ScreenSpot-Pro.	★★★★★		🇺🇸 Holo1.5-72B	63.2%
284	ScreenSpot-v2 O	Screen UI locators	Center accuracy on ScreenSpot-v2.	★★★★★		G UI-Venus 72B	95.3%
285	SEED-Bench-2-Plus o	Multimodal evaluation	SEED-Bench-2-Plus overall accuracy.	★★★★★		🇺🇸 Claude 3.7 Sonnet	72.9%
286	SEED-Bench-Img O	Multimodal image understanding	SEED-Bench image-only subset (SEED-Bench-Img).	★★★★★		G Bagel 14B	78.5%
287	Showdown O	GUI agents	Success rate on the Showdown UI interaction benchmark.	★★★★★		🇺🇸 Holo1.5-72B	76.8%
288	SIFO o	Instruction following	Single-turn instruction following benchmark.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Thinking	66.9%
289	SIFO Multiturn o	Instruction following	Multi-turn SIFO benchmark for sustained instruction adherence.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Thinking	60.3%
290	SimpleQA O	QA	Simple question answering benchmark.	★★★★★		🇨🇳 DeepSeek V3.2-Exp	97.1%
291	SimpleVQA o	General VQA	Lightweight visual question answering set with everyday scenes.	★★★★★		🇺🇸 Gemini 2.5 Pro	65.4%
292	SimpleVQA-DS o	General VQA	SimpleVQA variant curated by DeepSeek with everyday image question answering tasks.	★★★★★		🇨🇳 Seed1.5-VL-Thinking	61.3%
293	SocialIQA o	Social commonsense	Social interaction commonsense QA.	★★★★★		🇺🇸 Gemma 3 PT 27B	54.9%
294	Spider o	Text-to-SQL	Complex text-to-SQL benchmark over cross-domain databases.	★★★★★		🇺🇸 Llama 3 70B Base	67.1%
295	Spiral-Bench O	Safety / sycophancy	A LLM-judged benchmark measuring sycophancy and delusion reinforcement.	★★★★★		🇺🇸 GPT-5	87.0%
296	SQuAD v1.1 o	Reading comprehension	Extractive QA from Wikipedia articles.	★★★★★	566	🇺🇸 Llama 3.1 405B Base	89.3%
297	SUNRGBD o	3D scene understanding	SUN RGB-D benchmark for indoor scene understanding from RGB-D imagery.	★★★★★		🇺🇸 GPT-5 Mini Minimal	45.8%
298	SuperGPQA O	Graduate-level QA	Harder GPQA variant assessing advanced graduate-level reasoning.	★★★★★		🇨🇳 Qwen3-235B-A22B-Thinking-2507	64.9%
299	SWE-Bench o	Code repair	Supervised software engineering benchmark across many repos and issues.	★★★★★	3442	🇺🇸 GPT-5 Codex	74.5%
300	SWE-Bench Multilingual o	Code repair (multilingual)	Multilingual variant of SWE-Bench for issue fixing.	★★★★★		🇨🇳 DeepSeek V3.2-Exp	57.9%
301	SWE-Bench Pro (Public) o	Software engineering	Public subset of the SWE-Bench Pro benchmark for software-engineering agents.	★★★★★		🇺🇸 GPT-5	23.3%
302	SWE-Bench Verified O	Code repair	Verified subset of SWE-Bench for issue fixing.	★★★★★		🇺🇸 Claude Sonnet 4.5	77.2%
303	SWE-Dev o	Code repair	Software engineering development and bug fixing benchmark.	★★★★★		🇺🇸 Claude Sonnet 4	67.1%
304	SysBench o	System prompts	System prompt understanding and adherence benchmark.	★★★★★		🇺🇸 GPT-4.1	74.1%
305	τ²-Bench (airline) o	Industry QA (airline)	τ²-Bench airline domain evaluation.	★★★★★		🇺🇸 o3	64.8%
306	τ²-Bench (retail) o	Industry QA (retail)	τ²-Bench retail domain evaluation.	★★★★★		🇺🇸 Claude Opus 4.1	82.4%
307	τ²-Bench (telecom) O	Industry QA (telecom)	τ²-Bench telecom domain evaluation.	★★★★★		🇺🇸 GPT-5	96.7%
308	TAU1-Airline o	Agent tasks (airline)	Tool-augmented agent evaluation in airline scenarios (TAU1).	★★★★★		🇺🇸 Gemini-2.5-Flash Thinking	54.0%
309	TAU1-Retail o	Agent tasks (retail)	Tool-augmented agent evaluation in retail scenarios (TAU1).	★★★★★		🇨🇳 Qwen3-235B-A22B	71.3%
310	TAU2-Airline O	Agent tasks (airline)	Tool-augmented agent evaluation in airline scenarios (TAU2).	★★★★★		🇺🇸 Claude Sonnet 4.5	70.0%
311	TAU2-Retail O	Agent tasks (retail)	Tool-augmented agent evaluation in retail scenarios (TAU2).	★★★★★		🇺🇸 Claude Opus 4.1	86.8%
312	TAU2-Telecom O	Agent tasks (telecom)	Tool-augmented agent evaluation in telecom scenarios (TAU2).	★★★★★		🇺🇸 Claude Sonnet 4.5	98.0%
313	Terminal-Bench O	Agent terminal tasks	Command-line task completion benchmark for agents.	★★★★★	637	🇺🇸 Claude Sonnet 4.5 (Thinking)	61.3%
314	Terminal-Bench Hard O	Agent terminal tasks	Hard subset of Terminal-Bench command-line agent tasks.	★★★★★		🇺🇸 Grok 4	37.6%
315	TextVQA O	Text-based VQA	Visual question answering that requires reading text in images.	★★★★★		🇨🇳 Qwen2-VL-72B	85.5%
316	TreeBench o	Reasoning with tree structures	Evaluates hierarchical/tree-structured reasoning and planning capabilities in LLMs/VLMs.	★★★★★		🇨🇳 GLM-4.5V	50.1%
317	TriQA o	Knowledge QA	Triadic question answering benchmark evaluating world knowledge and reasoning.	★★★★★		🇫🇷 Mixtral 8x22B	82.2%
318	TriviaQA O	Open-domain QA	Open-domain question answering benchmark built from trivia and web evidence.	★★★★★		🇺🇸 Gemma 3 PT 27B	85.5%
319	TriviaQA-Wiki o	Open-domain QA	TriviaQA subset answering using Wikipedia evidence.	★★★★★		🇺🇸 Llama 3.1 405B Base	91.8%
320	TruthfulQA O	Truthfulness / hallucination	Measures whether a model imitates human falsehoods (truthfulness).	★★★★★		🇨🇳 Qwen2.5 32B Instruct	70.3%
321	TruthfulQA (DE) o	Truthfulness / hallucination (German)	German translation of the TruthfulQA benchmark.	★★★★★		🇺🇸 Llama 3.3 70B Instruct	0.2%
322	TydiQA o	Cross-lingual QA	Typologically diverse QA across languages.	★★★★★	313	🇺🇸 Llama 3.1 405B Base	34.3%
323	V* o	Multimodal reasoning	V* benchmark accuracy.	★★★★★		G MiMo-VL 7B-RL	81.7%
324	VCT O	Virology capability (protocol troubleshooting)	Virology Capabilities Test: a benchmark that measures an LLM's ability to troubleshoot complex virology laboratory protocols.	★★★★★		🇺🇸 o3	43.8%
325	VibeEval O	Aesthetic/visual quality	VLM aesthetic evaluation with GPT scores.	★★★★★		🇺🇸 Gemini 2.5 Pro	76.4%
326	Video-MME o	Video understanding (multimodal)	Multimodal evaluation of video understanding and reasoning.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Instruct	74.5%
327	VideoMME (w/o sub) O	Video understanding	Video understanding benchmark without subtitles.	★★★★★		🇺🇸 Gemini 2.5 Pro	85.1%
328	VideoMME (w/sub) o	Video understanding	Video understanding benchmark with subtitles.	★★★★★		🇨🇳 GLM-4.5V	80.7%
329	VideoMMMU O	Multimodal video understanding	Video-based extension of MMMU evaluating temporal multimodal reasoning and perception across disciplines.	★★★★★		🇺🇸 GPT-5	84.6%
330	VisualWebBench O	Web UI understanding	Average accuracy on VisualWebBench.	★★★★★		🇺🇸 Holo1.5-72B	83.8%
331	VisuLogic O	Visual logical reasoning	Logical reasoning and compositionality benchmark for visual-language models.	★★★★★		🇨🇳 Seed1.5-VL-Thinking	35.9%
332	VitaBench o	Industry QA	Industry-focused benchmark evaluating domain QA performance.	★★★★★		🇺🇸 o3	35.3%
333	VL-RewardBench o	Reward modeling (VL)	Reward alignment benchmark for VLMs.	★★★★★		🇺🇸 Claude 3.7 Sonnet	67.4%
334	VLMs are Biased o	Multimodal bias	Evaluates whether VLMs truly 'see' vs. relying on memorized knowledge; measures bias toward non-visual priors.	★★★★★	90	🇺🇸 o4 mini	20.2%
335	VLMs are Blind O	Visual grounding robustness	Evaluates failure modes of VLMs in grounding and perception tasks.	★★★★★		G MiMo-VL 7B-RL	79.4%
336	VoiceBench AdvBench o	VoiceBench	VoiceBench adversarial safety evaluation.	★★★★★		🇨🇳 Qwen3-Omni-30B-A3B-Thinking	99.4%
337	VoiceBench AlpacaEval o	VoiceBench	VoiceBench evaluation on AlpacaEval instructions.	★★★★★		🇨🇳 Qwen3-Omni-Flash-Thinking	96.8%
338	VoiceBench BBH o	VoiceBench	VoiceBench evaluation on Big-Bench Hard prompts.	★★★★★		🇺🇸 Gemini 2.5 Pro	92.6%
339	VoiceBench CommonEval o	VoiceBench	VoiceBench evaluation on CommonEval.	★★★★★		🇨🇳 Qwen3-Omni-Flash-Instruct	91.0%
340	VoiceBench IFEval o	VoiceBench	VoiceBench instruction-following evaluation (IFEval).	★★★★★		🇺🇸 Gemini 2.5 Pro	85.7%
341	MMAU v05.15.25 o	Audio reasoning	Audio reasoning benchmark MMAU v05.15.25.	★★★★★		🇨🇳 Qwen3-Omni-Flash-Instruct	77.6%
342	VoiceBench MMSU o	VoiceBench	VoiceBench MMSU benchmark (voice modality).	★★★★★		🇨🇳 Qwen3-Omni-Flash-Thinking	84.3%
343	VoiceBench MMSU (Audio) o	Audio reasoning	Audio reasoning MMSU results.	★★★★★		🇺🇸 Gemini 2.5 Pro	77.7%
344	VoiceBench OpenBookQA o	VoiceBench	VoiceBench results on OpenBookQA prompts.	★★★★★		🇨🇳 Qwen3-Omni-Flash-Thinking	95.0%
345	VoiceBench Overall o	VoiceBench	Overall VoiceBench aggregate score.	★★★★★		🇺🇸 Gemini 2.5 Pro	89.6%
346	VoiceBench SD-QA o	VoiceBench	VoiceBench Spoken Dialogue QA results.	★★★★★		🇺🇸 Gemini 2.5 Pro	90.1%
347	VoiceBench WildVoice o	VoiceBench	VoiceBench evaluation on WildVoice dataset.	★★★★★		🇺🇸 Gemini 2.5 Pro	93.4%
348	VQAv2 O	Visual question answering	Standard Visual Question Answering v2 benchmark on natural images.	★★★★★		🇺🇸 Molmo-72B	86.5%
349	VSI-Bench o	Spatial intelligence	Visual spatial intelligence benchmark covering 3D reasoning and spatial inference tasks.	★★★★★		🇨🇳 Qwen3-VL-30B-A3B Instruct	63.2%
350	WebClick O	GUI agents	Task success on the WebClick UI agent benchmark.	★★★★★		🇺🇸 Claude Sonnet 4	93.0%
351	WebDev Arena O	Web development agents	Arena evaluation for autonomous web development agents.	★★★★★		🇺🇸 GPT-5	1483
352	WebQuest-MultiQA o	Web agents	Multi-question web search and interaction tasks.	★★★★★		🇨🇳 GLM-4.5V	60.6%
353	WebQuest-SingleQA o	Web agents	Single-question web search and interaction tasks.	★★★★★		🇨🇳 GLM-4.5V	76.9%
354	WebSrc O	Web QA	Webpage question answering (SQuAD F1).	★★★★★		🇺🇸 Holo1.5-72B	97.2%
355	WebVoyager2 o	Web agents	Web navigation and interaction tasks for LLM agents (v2).	★★★★★		🇨🇳 GLM-4.5V	84.4%
356	WebWalkerQA o	Web agents	WebWalker tasks evaluating autonomous browsing question answering performance.	★★★★★		🇨🇳 Tongyi DeepResearch	72.2%
357	WeMath o	Math reasoning	Math reasoning benchmark spanning diverse curricula and difficulty levels.	★★★★★		🇨🇳 GLM-4.5V	68.8%
358	WildBench V2 o	Instruction following	WildBench V2 human preference benchmark for instruction following and helpfulness.	★★★★★		🇫🇷 Mistral Small 3.2 24B Instruct	65.3%
359	Winogender o	Gender bias (coreference)	Coreference resolution dataset for measuring gender bias.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	67.9%
360	WinoGrande O	Coreference reasoning	Large-scale adversarial Winograd Schema-style pronoun resolution.	★★★★★	99	🇺🇸 Llama 3.1 405B Base	86.7%
361	WinoGrande (DE) o	Coreference reasoning (German)	German translation of the WinoGrande pronoun resolution benchmark.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	0.8%
362	WMT16 En–De o	Machine translation	WMT16 English–German translation benchmark (news).	★★★★★		🇺🇸 Llama 3.3 70B Instruct	38.8%
363	WMT16 En–De (Instruct) o	Machine translation	Instruction-tuned evaluation on the WMT16 English–German translation set.	★★★★★		🇺🇸 Llama 3.3 70B Instruct	37.9%
364	WritingBench O	Writing quality	General-purpose writing quality benchmark.	★★★★★		🇨🇳 Qwen3-235B-A22B-Thinking-2507	88.3%
365	WSC o	Coreference reasoning	Classic Winograd Schema Challenge measuring commonsense coreference.	★★★★★		G Pythia 410M	47.1%
366	xBench-DeepSearch o	Agentic research	Evaluates multi-hop deep research workflows on xBench DeepSearch tasks.	★★★★★		🇨🇳 Tongyi DeepResearch	75.0%
367	ZebraLogic O	Logical reasoning	Logical reasoning benchmark assessing complex pattern and rule inference.	★★★★★		🇨🇳 Qwen3 MoE-2507	94.2%
368	ZeroBench O	Zero-shot generalization	Evaluates zero-shot performance across diverse tasks without task-specific finetuning.	★★★★★		🇨🇳 GLM-4.5V	23.4%
369	ZeroBench (sub) O	Zero-shot generalization	Subset of ZeroBench targeting harder zero-shot reasoning cases.	★★★★★		🇺🇸 Gemini 2.5 Pro	33.8%
370	ZeroSCROLLS MuSiQue o	Long-context reasoning	ZeroSCROLLS split derived from MuSiQue multi-hop QA.	★★★★★		🇺🇸 Llama 3.3 70B Instruct	0.5%
371	ZeroSCROLLS SpaceDigest o	Long-context summarization	ZeroSCROLLS SpaceDigest extractive summarization task.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	0.8%
372	ZeroSCROLLS SQuALITY o	Long-context summarization	ZeroSCROLLS split based on the SQuALITY long-form summarization benchmark.	★★★★★		🇺🇸 Llama-3_1-70B-TFree-HAT-SFT	0.2%