| ProverBench | DeepSeek | DeepSeek-Prover-V2-671B (CoT) |
| GSM-Symbolic | Apple | o1-preview |
| MATH | UC Berkeley | GPT-2 1.5B |
| ProcessBench | Alibaba Inc. | o1-mini |
| Mathador-LM | ISTA | Llama-3-70B |
| IMO-Bench | Google DeepMind | Gemini Deep Think (IMO Gold) |
| FrontierMath | Harvard University | o1-preview |
| MiniF2F | University of Pittsburgh | GPT-f |
| OlympiadBench | Beihang University | GPT-4 |
| MathVista | University of Washington | InternVL2-Pro |
| MATH-Perturb | Google | o1-mini |
| REASONEVAL | Shanghai Artificial Intelligence Laboratory | REASONEVAL (Llemma-34B) |
| Omni-MATH | University of Waterloo | OpenAI o1-mini |
| OMEGA | University of Washington | DeepSeek-R1 |
| MathConstruct | ETH Zurich | GPT-5 |
| FormalMATH | UCLA | DeepSeek-V2-671B |
| QRDATA | UCLA | GPT-4 |
| ME2 | Yonsei University | Gemini 2.0 Flash |
| Math Reasoning Benchmark | Technical University of Munich | GPT-4o |
| GSM-Infinite | Meta | DeepSeek-R1 |
| OptiBench | Huawei Noah’s Ark Lab | GPT-4o |
| MathBookEval | Tsinghua University | Qwen2.5-VL-72B |
| IneqMath | UC Berkeley | GPT-5 (medium, 30K) |
| GSM-IC | Google Research | LTM + SC (code-davinci-002) |
| MGSM | Stanford University | PaLM-540B |
| GSM-Plus | Tencent AI Lab | Human |
| REST | Shanghai Artificial Intelligence Laboratory | Qwen-QwQ-32B |
| NUPA | Peking University | GPT-4o |
| LiveAoPSBench | University of British Columbia | DeepSeek-R1-Distill-Qwen-32B |
| FinanceMATH | New York University | Human Expert (Open-book) |
| Mamo | Shenzhen Research Institute of Big Data | Llama-3.1-405B-instruct |
| OlymMATH | BAAI | Gemini 2.5 Pro Exp 0325 |
| CHASE | Université de Montréal | Gemini-1.5-Pro |
| PutnamBench | UT Austin | Hilbert |
| MWPBENCH | Shenzhen Research Institute of Big Data | GPT-4 |
| MATHCHECK | University of Macau | O1-preview |
| ArithmAttack | University of Bonn | Llama3 |
| Math-VR | CUHK | Qwen3-VL-235B-A22B-Thinking |
| SciBench | University of Washington | GPT-4-Turbo |
| GeomVerse | Google DeepMind | Human |
| MathCanvas-Bench | The Chinese University of Hong Kong | Gemini-2.5-Pro |
| MCLM | Yonsei University | o3-Mini |
| We-Math | Huazhong University of Science and Technology | Gemini-2.5-Pro |
| GeoTrust | Shanghai Artificial Intelligence Laboratory | OpenAI-o3 |
| GSM1k | Scale AI | gpt-4o |
| MathChat | University of Notre Dame | Mistral-MathChat |
| DynaMath | University of Illinois at Urbana-Champaign | Zero-shot Claude-3.5 |
| Cultural GSM8K | Griffith University | Claude 3.5 Sonnet |
| GeoSense | Alibaba Group | Gemini 1.5 Pro |
| CMM-Math | East China Normal University | GPT-4o |
| Hard2Verify | Salesforce AI Research | GPT-5 |
| MathArena | ETH Zurich | GPT-5 (HIGH) |
| DocMath-Eval | New York University | Gemini-1.5-Pro |
| COUNTERMATH | Sun Yat-Sen University | Deepseek-R1 |
| FATE-M | National University of Singapore | REAL-Prover |
| Geometry3K | UCLA | Inter-GPS |
| HARDMath | Harvard University | o1-mini |
| MathOdyssey | Johns Hopkins University | OpenAI o1 |
| DeepMath-Creative | Tongji University | GPT O3-mini |
| REWARDMATH | Yonsei University | Internlm2-7b-reward |
| RealMath | ETH Zurich | o3 |
| ErrorRadar | HKUST(GZ) | Human |
| CHAMP | Boston University | GPT-4 Turbo |
| U-MATH | Stevens Institute of Technology | Gemini-1.5-Pro |
| Numberland | SLAC National Accelerator Laboratory | ChatGPT o1 |
| MM-MATH | Tsinghua University | Human |
| REASONZOO | Shanghai Artificial Intelligence Laboratory | Qwen3-235B-A22B |
| UniGeo | UCLA | Geoformer + Pretraining |
| SOLIDGEO | Chinese Academy of Sciences | Human |
| CombiBench | Chinese Academy of Sciences | Kimina-Prover Preview |
| Libra Bench | Meituan | Libra-RM-32B-MATH |
| Scheherazade | Yale University | o1-preview |
| TABMWP | UCLA | Docugami-MATATA-8B |
| Advanced Reasoning Benchmark | Université de Montréal | gpt-4-0314 |
| Geoint | Chinese Academy of Sciences | Geoint-R1 |
| ModelingBench | University of Illinois at Urbana-Champaign | Human Expert |
| AMO-Bench | Chinese Academy of Sciences | GPT-5-Thinking (High) |
| SVAMP | Microsoft | Graph2Tree-R |
| FERMAT | Chennai Mathematical Institute | GEMINI-1.5-PRO |
| LeanGeo | University of Toronto | Gemini 2.5 Pro |
| GeoGramBench | Li Auto | Qwen3-235B-Thinking-2507 |
| GeoEval | Chinese Academy of Sciences | WizardMath-70B |
| MATP-BENCH | HKUST | OpenAI-o1 |
| Ineq-Comp | Tsinghua University | DeepSeek-Prover-V2-7B |
| MathReal | Beihang University | DeepSeek-R1 |
| GeomRel | Shanghai Jiao Tong University | GPT-4o |
| PROOFBENCH | Google DeepMind | O3 |
| GPSM4K | NUS | GPT-4 |
| RV-BENCH | University of Electronic Science and Technology of China | o3-mini |
| FinanceReasoning | Beijing University of Posts and Telecommunications | OpenAI o1 |
| StepMathBench | Tsinghua University | o1-mini |
| UGMathBench | HKUST | OpenAI-o1-mini-2024-09-12 |
| ReliableMath | Huawei Noah’s Ark Lab | DeepSeek-R1 |
| MathGAP | Max Planck Institute for Intelligent Systems | DeepSeek-R1 |
| HARP | Anthropic | o1 mini |
| VisionGraph | Sun Yat-Sen University | GPT-4V (DPR) w/ Python |
| Extended Grade-School Math | University of Science and Technology of China | Claude-3-opus |
| Formal Problem-Solving Benchmarks | Shanghai Artificial Intelligence Laboratory | InternLM2.5-StepProver |
| ConsistencyCheck | Alibaba Group | ReForm-32B |
| UTMath | South China University of Technology | o1-mini |
| MMATH | Renmin University of China | o3-mini |
| CMATH | Xiaomi | GPT-4 |
| FineMath | Tianjin University | GPT-4 |
| PutnamGAP | University of Illinois at Urbana-Champaign | gpt-4o-mini |
| Unreasonable Math Problems | Peking University | Grok3-Reasoning |
| TriMaster100 | Technical University of Munich | SSC-CoT (GPT-3.5) |
| MathHay | Singapore Management University | Gemini-1.5-Pro-002 |
| FINEREASON | Nanyang Technological University | o1 |
| TreeCut | Columbia University | o3-mini |
| PMC | Nanjing University | Qwen2.5 3B |
| MATH 401 | Alibaba Group | gpt-4 |
| SceMQA | University of Notre Dame | GPT4-V (Zero-shot) |
| FormulaReasoning | State Key Laboratory for Novel Software Technology, Nanjing University | Human |
| EasyMath | Assistantslab | AceMath-1.5B |
| GSM-Agent | UC Berkeley | o3 |
| AgentCoMa | Imperial College London | Llama3.3 70B Instruct |
| MATH-Beyond | University of Tübingen | Qwen3-8B |
| JEEBENCH | IIT Delhi | GPT-4 + CoT + SC@8 |
| BrokenMath | ETH Zurich | GPT-5 |
| CogMath | University of Science and Technology of China | DeepSeek-R1 |
| MMSciBench | Fudan University | Gemini 1.5 Pro 002 |
| LILA | Harvard University | Codex (code-davinci-002) |
| Invalsi Benchmarks | ISTI-CNR | llama 3.1 70b instruct |
| MME-SCI | Shanghai Jiao Tong University | Doubao-Seed-1.6 |
| ConceptMath | Shanghai Artificial Intelligence Laboratory | GPT-4 |
| CREATIVEMATH | Pennsylvania State University | Gemini-1.5-Pro |
| GSM-MC | Shanghai Artificial Intelligence Laboratory | Qwen3-8B |
| NumGLUE | Microsoft | Human |
| Spoken-MQA | Institute for Infocomm Research (I2R), A*STAR | Whisper-Qwen2.5-Math-7B-Instruct |
| StatEval | Shanghai University of Finance and Economics | GPT-5 |
| SMART | Beijing Normal University | o3 |
| PATCH! | Utrecht University | GPT-4V |
| RIMO | Hong Kong University of Science and Technology (Guangzhou) | DeepSeek-R1-671B |
| MathMist | University of Virginia | GPT-OSS 20B |
| ASDiv | Academia Sinica | GTS |
| Basic Math Benchmark | Virginia Tech | Llama-2-13b |
| EvolMathEval | Sun Yat-Sen University | DeepSeek-R1 |
| TRIGO | Huawei Noah’s Ark Lab | GPT-2_L-PACT-E |
| FMC | University of Washington | DEEPSEEK-R1 |
| VisAidMath | Alibaba Group | GPT-4V |
| KMath | Kuaishou Technology Inc. | GPT-4 |
| MorphoBench | Chinese Academy of Sciences | o3 |
| SciDA | ByteDance | Gemini-2.5-pro.preview.0506.google.ci |
| IsarStep | University of Toronto | HAT |
| BeyondX | UCLA | GPT-4 |
| SuperCLUE-Math6 | SuperCLUE team | GPT-4-1106-Preview |
| AVI-MATH | Shanghai Jiao Tong University | AVI-Math (Ours) |
| MATH-Struct | Shanghai Artificial Intelligence Laboratory | Qwen-2-7B |
| FAULTYMATH | Pennsylvania State University | Gemini-1.5-Pro |
| ASyMOB | Technion
Israel Institute of Technology | Gemini-2.5 Flash (no code) |
| Putnam-AXIOM | Stanford University | o1-preview |
| Kangaroo Language Test | KTH Royal Institute of Technology | Gemini 2.0 Flash |
| Mathematical Topics Tree | Carnegie Mellon University | o1-mini |
| MMTutorBench | University of Notre Dame | Gemini-2.5-Pro |
| AI4Math | UNAM | o3 mini |
| Combi-Puzzles | University of Cambridge | GPT-4 |
| CM17K | UCLA | NS-Solver |
| UAV-Math-Bench | University of Stuttgart | o1 |
| FATE | Westlake University | o3 |
| RoMath | University of Bucharest | Mathstral-7b-v0.1 (0-shot) |
| NOAHQA | University of Electronic Science and Technology of China | Human |
| DOoM | HSE University | Gemini 2.5 Pro |
| SKYLENAGE | Alibaba Group | GPT-5-20250807 |
| AlgGeoTest | Peking University | Gemini 2.5 Pro |
| IntegralBench | Tongji University | Qwen3-235B-A22B |
| InterMWP | Sun Yat-Sen University | LogicSolver |
| Guji_MATH | Nanjing Agricultural University | DeepSeek R1 |
| Machine Number Sense | UCLA | ResNet |
| ORCA | Leibniz University Hannover | Gemini 2.5 Flash |
| MAVEN | CoreThink AI | CoreThink/openai/gpt-oss-120b |
| EffiReason-Bench | Nanyang Technological University | TokenSkip |
| DynaSolidGeo | Peking University | Qwen3-VL-30B-A3B-Thinking |
| MatheMagic | University of Edinburgh | GPT-5 |
| EQUATE | Carnegie Mellon University | Q-REAS |
| MathRobust-LV | Got It Education | GPT-5 |
| MathBode | Columbia University | DeepSeek V3.1 |
| ExtremBench | Fudan University | Qwen3-4B-Thinking-2507 |
| GanitBench | Vellore Institute of Technology | GPT-4o mini (Zero-shot CoT) |
| PuzzleClone | HiThink Research | ChatGPT-o3 |
| StreetMath | University of Waterloo | Qwen3-4B-Instruct |
| BanglaMATH | Santa Fe Institute | DeepSeek-V3 |
| MathSticks | Chinese Academy of Sciences | Human |