Leaderboard
MathVista
mathvistaRecord
- Category
- multimodal
- Scored models
- 96
Scores
| Model | Score |
|---|---|
| moonshotai/Kimi-VL-A3B-Instruct | 80.1 |
| nvidia/NVIDIA-Nemotron-Nano-12B-v2-VL-BF16 | 76.9 |
| Qwen/Qwen2.5-VL-72B-Instruct | 75.3 |
| microsoft/Phi-4-reasoning | 75.2 |
| Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen/Qwen2.5-VL-32B-Instruct | 74.7 |
| Qwen/Qwen2.5-VL-7B-Instruct | 73.8 |
| Qwen/Qwen2.5-VL-3B-Instruct | 73.8 |
| nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| Qwen/Qwen2-VL-72B | 71.4 |
| inclusionAI/M2-Reasoning | 70.5 |
| Qwen/Qwen2-VL-72B-Instruct | 70.19 |
| Qwen/Qwen2-VL-72B-Instruct | 70.19 |
| Qwen/Qwen2-VL-72B-Instruct | 70.19 |
| mistralai/Mistral-Small-3.1-24B-Base-2503 | 68.91 |
| mistralai/Mistral-Small-3.1-24B-Base-2503 | 68.91 |
Remaining fields
Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.
- category
- multimodal
- id
- mathvista
- name
- MathVista
- schema version
- local-ai-registry/v1
rows
| variant | conf | context | model id | org | rank | root | score |
|---|---|---|---|---|---|---|---|
| Kimi-VL-A3B-Thinking-2506 | medium | MathVista\_MINI (Pass@1) | kimi-vl-a3b-instruct | moonshotai | 1 | moonshotai/Kimi-VL-A3B-Instruct | 80.1 |
| NVIDIA-Nemotron-Nano-12B-v2-VL-BF16 | medium | MathVista* | Unknown | nvidia | 2 | nvidia/NVIDIA-Nemotron-Nano-12B-v2-VL-BF16 | 76.9 |
| Qwen2.5-VL-72B-Instruct-AWQ | high | Unknown | qwen2-5-vl-72b-instruct | Qwen | 3 | Qwen/Qwen2.5-VL-72B-Instruct | 75.3 |
| Phi-4-reasoning-vision-15B | medium | MathVista_MINI | phi-4-reasoning | microsoft | 4 | microsoft/Phi-4-reasoning | 75.2 |
| Qwen2.5-VL-32B-Instruct | medium | MathVista | qwen2-5-vl-32b-instruct | Qwen | 5 | Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen2.5-VL-32B-Instruct-GGUF | medium | MathVista | qwen2-5-vl-32b-instruct | unsloth | 6 | Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen2.5-VL-32B-Instruct-bnb-4bit | medium | MathVista | qwen2-5-vl-32b-instruct | unsloth | 7 | Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen2.5-VL-32B-Instruct-GGUF | medium | MathVista | qwen2-5-vl-32b-instruct | Mungert | 8 | Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen2.5-VL-32B-Instruct-unsloth-bnb-4bit | medium | MathVista | qwen2-5-vl-32b-instruct | unsloth | 9 | Qwen/Qwen2.5-VL-32B-Instruct | 74.8 |
| Qwen2.5-VL-32B-Instruct-AWQ | high | Unknown | qwen2-5-vl-32b-instruct | Qwen | 10 | Qwen/Qwen2.5-VL-32B-Instruct | 74.7 |
| Qwen2.5-VL-7B-Instruct-AWQ | high | Unknown | qwen2-5-vl-7b-instruct | Qwen | 11 | Qwen/Qwen2.5-VL-7B-Instruct | 73.8 |
| Qwen2.5-VL-3B-Instruct-AWQ | high | Unknown | qwen2-5-vl-3b-instruct | Qwen | 12 | Qwen/Qwen2.5-VL-3B-Instruct | 73.8 |
| Nemotron-3-Nano-Omni-30B-A3B-Reasoning-NVFP4 | medium | MathVista_MINI | nemotron-3-nano-omni-30b-a3b-reasoning-bf16 | nvidia | 13 | nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| Nemotron-3-Nano-Omni-30B-A3B-Reasoning-FP8 | medium | MathVista_MINI | nemotron-3-nano-omni-30b-a3b-reasoning-bf16 | nvidia | 14 | nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | medium | MathVista_MINI | nemotron-3-nano-omni-30b-a3b-reasoning-bf16 | nvidia | 15 | nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| NVIDIA-Nemotron-3-Nano-Omni-30B-A3B-Reasoning-GGUF | medium | MathVista_MINI | nemotron-3-nano-omni-30b-a3b-reasoning-bf16 | unsloth | 16 | nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| NVIDIA-Nemotron-3-Nano-Omni-30B-A3B-Reasoning | medium | MathVista_MINI | nemotron-3-nano-omni-30b-a3b-reasoning-bf16 | unsloth | 17 | nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 | 71.9 |
| QVQ-72B-Preview | medium | MathVista(mini) | Unknown | Qwen | 18 | Qwen/Qwen2-VL-72B | 71.4 |
| M2-Reasoning | low | Unknown | Unknown | inclusionAI | 19 | inclusionAI/M2-Reasoning | 70.5 |
| Qwen2-VL-72B-Instruct-AWQ | high | Unknown | qwen2-vl-72b-instruct | Qwen | 20 | Qwen/Qwen2-VL-72B-Instruct | 70.19 |
| Qwen2-VL-72B-Instruct-GPTQ-Int4 | high | Unknown | qwen2-vl-72b-instruct | Qwen | 21 | Qwen/Qwen2-VL-72B-Instruct | 70.19 |
| Qwen2-VL-72B-Instruct-GPTQ-Int8 | high | Unknown | qwen2-vl-72b-instruct | Qwen | 22 | Qwen/Qwen2-VL-72B-Instruct | 70.19 |
| Mistral-Small-3.1-24B-Instruct-2503 | low | Unknown | mistral-small-3-1-24b-base-2503 | mistralai | 23 | mistralai/Mistral-Small-3.1-24B-Base-2503 | 68.91 |
| Mistral-Small-3.2-24B-Instruct-2506 | low | Unknown | mistral-small-3-1-24b-base-2503 | mistralai | 24 | mistralai/Mistral-Small-3.1-24B-Base-2503 | 68.91 |
| Mistral-Small-3.2-24B-Instruct-2506-GGUF | low | Unknown | mistral-small-3-1-24b-base-2503 | unsloth | 25 | mistralai/Mistral-Small-3.2-24B-Instruct-2506 | 68.91 |
| Mistral-Small-3.2-24B-Instruct-2506-bnb-4bit | low | Unknown | mistral-small-3-1-24b-base-2503 | unsloth | 26 | mistralai/Mistral-Small-3.2-24B-Instruct-2506 | 68.91 |
| Mistral-Small-3.2-24B-Instruct-2506 | low | Unknown | mistral-small-3-1-24b-base-2503 | unsloth | 27 | mistralai/Mistral-Small-3.2-24B-Instruct-2506 | 68.91 |
| Mistral-Small-3.1-24B-Instruct-2503-GGUF | low | Unknown | mistral-small-3-1-24b-instruct-2503 | unsloth | 28 | mistralai/Mistral-Small-3.1-24B-Instruct-2503 | 68.91 |
| Mistral-Small-3.2-24B-Instruct-2506-unsloth-bnb-4bit | low | Unknown | mistral-small-3-1-24b-base-2503 | unsloth | 29 | mistralai/Mistral-Small-3.2-24B-Instruct-2506 | 68.91 |
| Mistral-Small-3.1-24B-Instruct-2503-bnb-4bit | low | Unknown | mistral-small-3-1-24b-instruct-2503 | unsloth | 30 | mistralai/Mistral-Small-3.1-24B-Instruct-2503 | 68.91 |
| Mistral-Small-3.1-24B-Instruct-2503-unsloth-bnb-4bit | low | Unknown | mistral-small-3-1-24b-instruct-2503 | unsloth | 31 | mistralai/Mistral-Small-3.1-24B-Instruct-2503 | 68.91 |
| Mistral-Small-3.1-24B-Instruct-2503 | low | Unknown | mistral-small-3-1-24b-instruct-2503 | unsloth | 32 | mistralai/Mistral-Small-3.1-24B-Instruct-2503 | 68.91 |
| Mistral-Small-3.1-24B-Instruct-2503-GGUF | low | Unknown | mistral-small-3-1-24b-base-2503 | Mungert | 33 | mistralai/Mistral-Small-3.1-24B-Base-2503 | 68.91 |
| Mistral-Small-3.2-24B-Instruct-2506-GGUF | low | Unknown | mistral-small-3-1-24b-base-2503 | Mungert | 34 | mistralai/Mistral-Small-3.1-24B-Base-2503 | 68.91 |
| Mistral-Small-3.2-24B-Instruct-2506-FP8 | low | Unknown | mistral-small-3-1-24b-base-2503 | unsloth | 35 | mistralai/Mistral-Small-3.2-24B-Instruct-2506 | 68.91 |
| Ming-Lite-Omni | medium | MathVista | ling-lite | inclusionAI | 36 | inclusionAI/Ling-lite | 68.1 |
| Qwen2.5-Omni-3B | medium | MathVista<sub>testmini</sub> | qwen2-5-omni-3b | Qwen | 37 | Qwen/Qwen2.5-Omni-3B | 67.9 |
| Qwen2.5-Omni-7B | medium | MathVista<sub>testmini</sub> | qwen2-5-omni-7b | Qwen | 38 | Qwen/Qwen2.5-Omni-7B | 67.9 |
| Qwen2.5-Omni-7B-GGUF | medium | MathVista<sub>testmini</sub> | qwen2-5-omni-7b | unsloth | 39 | Qwen/Qwen2.5-Omni-7B | 67.9 |
| Qwen2.5-Omni-3B-GGUF | medium | MathVista<sub>testmini</sub> | qwen2-5-omni-3b | unsloth | 40 | Qwen/Qwen2.5-Omni-3B | 67.9 |
| Qwen2.5-Omni-3B-GGUF | medium | MathVista<sub>testmini</sub> | qwen2-5-omni-3b | Mungert | 41 | Qwen/Qwen2.5-Omni-3B | 67.9 |
| Qwen2.5-Omni-7B | medium | MathVista<sub>testmini</sub> | qwen2-5-omni-7b | unsloth | 42 | Qwen/Qwen2.5-Omni-7B | 67.9 |
| Qwen2.5-Omni-7B-GGUF | medium | MathVista<sub>testmini</sub> | qwen2-5-omni-7b | Mungert | 43 | Qwen/Qwen2.5-Omni-7B | 67.9 |
| Qwen2.5-Omni-3B | medium | MathVista<sub>testmini</sub> | qwen2-5-omni-3b | unsloth | 44 | Qwen/Qwen2.5-Omni-3B | 67.9 |
| Pixtral-Large-Instruct-2411 | low | MathVista (CoT) | MMMU (CoT) | ChartQA (CoT) | Unknown | mistralai | 45 | mistralai/Pixtral-Large-Instruct-2411 | 67.8 |
| NVLM-D-72B | low | Unknown | nvlm-d-72b | nvidia | 46 | nvidia/NVLM-D-72B | 65.2 |
| Qwen2.5-VL-72B-Instruct | medium | MathVista_MINI | qwen2-5-vl-72b-instruct | Qwen | 47 | Qwen/Qwen2.5-VL-72B-Instruct | 63.8 |
| Kimi-VL-A3B-Instruct | medium | MathVista (Pass@1) | Unknown | moonshotai | 48 | moonshotai/Moonlight-16B-A3B | 63.8 |
| Kimi-VL-A3B-Thinking | medium | MathVista (mini) (Pass@1) | kimi-vl-a3b-instruct | moonshotai | 49 | moonshotai/Kimi-VL-A3B-Instruct | 63.8 |
| Eagle2.5-8B | medium | MathVista<sub>testmini</sub> | qwen2-5-7b-instruct | nvidia | 50 | Qwen/Qwen2.5-7B-Instruct | 63.8 |
| Kimi-VL-A3B-Thinking-2506 | medium | MathVista_MINI (Pass@1) | kimi-vl-a3b-instruct | moonshotai | 51 | moonshotai/Kimi-VL-A3B-Instruct | 63.8 |
| Qwen2.5-VL-72B-Instruct-GGUF | medium | MathVista_MINI | qwen2-5-vl-72b-instruct | unsloth | 52 | Qwen/Qwen2.5-VL-72B-Instruct | 63.8 |
| Qwen2.5-VL-72B-Instruct-GGUF | medium | MathVista_MINI | qwen2-5-vl-72b-instruct | Mungert | 53 | Qwen/Qwen2.5-VL-72B-Instruct | 63.8 |
| Qwen2.5-VL-72B-Instruct-bnb-4bit | medium | MathVista_MINI | qwen2-5-vl-72b-instruct | unsloth | 54 | Qwen/Qwen2.5-VL-72B-Instruct | 63.8 |
| Qwen2.5-VL-72B-Instruct-unsloth-bnb-4bit | medium | MathVista_MINI | qwen2-5-vl-72b-instruct | unsloth | 55 | Qwen/Qwen2.5-VL-72B-Instruct | 63.8 |
| Phi-4-multimodal-instruct | medium | MathVista (testmini) | phi-4-multimodal-instruct | microsoft | 56 | microsoft/Phi-4-multimodal-instruct | 62.4 |
| Phi-4-multimodal-instruct | medium | MathVista (testmini) | phi-4-multimodal-instruct | Lexius | 57 | microsoft/Phi-4-multimodal-instruct | 62.4 |
| MiniMax-VL-01 | medium | MathVista<sup>*</sup> | minimax-vl-01 | MiniMaxAI | 58 | MiniMaxAI/MiniMax-VL-01 | 62.1 |
| Qwen2.5-VL-3B-Instruct | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-3b-instruct | Qwen | 59 | Qwen/Qwen2.5-VL-3B-Instruct | 60.5 |
| Qwen2.5-VL-3B-Instruct-unsloth-bnb-4bit | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-3b-instruct | unsloth | 60 | Qwen/Qwen2.5-VL-3B-Instruct | 60.5 |
| Qwen2.5-VL-3B-Instruct-GGUF | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-3b-instruct | unsloth | 61 | Qwen/Qwen2.5-VL-3B-Instruct | 60.5 |
| Qwen2.5-VL-3B-Instruct-GGUF | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-3b-instruct | Mungert | 62 | Qwen/Qwen2.5-VL-3B-Instruct | 60.5 |
| Qwen2.5-VL-3B-Instruct-bnb-4bit | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-3b-instruct | unsloth | 63 | Qwen/Qwen2.5-VL-3B-Instruct | 60.5 |
| Qwen2.5-VL-3B-Instruct | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-3b-instruct | unsloth | 64 | Qwen/Qwen2.5-VL-3B-Instruct | 60.5 |
| Qwen2.5-VL-7B-Instruct | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-7b-instruct | Qwen | 65 | Qwen/Qwen2.5-VL-7B-Instruct | 58.3 |
| Qwen2-VL-7B-Instruct | medium | MathVista<sub>testmini</sub> | qwen2-vl-7b | Qwen | 66 | Qwen/Qwen2-VL-7B | 58.3 |
| Qwen2.5-VL-7B-Instruct-GGUF | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-7b-instruct | unsloth | 67 | Qwen/Qwen2.5-VL-7B-Instruct | 58.3 |
| Qwen2.5-VL-7B-Instruct | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-7b-instruct | unsloth | 68 | Qwen/Qwen2.5-VL-7B-Instruct | 58.3 |
| Qwen2.5-VL-7B-Instruct-unsloth-bnb-4bit | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-7b-instruct | unsloth | 69 | Qwen/Qwen2.5-VL-7B-Instruct | 58.3 |
| Qwen2.5-VL-7B-Instruct-bnb-4bit | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-7b-instruct | unsloth | 70 | Qwen/Qwen2.5-VL-7B-Instruct | 58.3 |
| Qwen2-VL-7B-Instruct-unsloth-bnb-4bit | medium | MathVista<sub>testmini</sub> | qwen2-vl-7b-instruct | unsloth | 71 | Qwen/Qwen2-VL-7B-Instruct | 58.3 |
| Qwen2.5-VL-7B-Instruct-GGUF | medium | MathVista<sub>testmini</sub> | qwen2-5-vl-7b-instruct | Mungert | 72 | Qwen/Qwen2.5-VL-7B-Instruct | 58.3 |
| Qwen2-VL-7B-Instruct-bnb-4bit | medium | MathVista<sub>testmini</sub> | qwen2-vl-7b-instruct | unsloth | 73 | Qwen/Qwen2-VL-7B-Instruct | 58.3 |
| Qwen2-VL-7B-Instruct | medium | MathVista<sub>testmini</sub> | qwen2-vl-7b-instruct | unsloth | 74 | Qwen/Qwen2-VL-7B-Instruct | 58.3 |
| Qwen2-VL-7B-Instruct-AWQ | high | Unknown | qwen2-vl-7b-instruct | Qwen | 75 | Qwen/Qwen2-VL-7B-Instruct | 58.2 |
| Qwen2-VL-7B-Instruct-GPTQ-Int8 | high | Unknown | qwen2-vl-7b-instruct | Qwen | 76 | Qwen/Qwen2-VL-7B-Instruct | 58.2 |
| Qwen2-VL-7B-Instruct-GPTQ-Int4 | high | Unknown | qwen2-vl-7b-instruct | Qwen | 77 | Qwen/Qwen2-VL-7B-Instruct | 58.2 |
| Pixtral-12B-2409 | medium | **Mathvista** *(CoT)* | Unknown | mistralai | 78 | mistralai/Pixtral-12B-Base-2409 | 54.4 |
| gemma-3-270m | medium | [MathVista][] (testmini) | gemma-3-270m | 79 | google/gemma-3-270m | 50 | |
| gemma-3-270m-it | medium | [MathVista][] (testmini) | gemma-3-270m | 80 | google/gemma-3-270m | 50 | |
| gemma-3-270m-it-GGUF | medium | [MathVista][] (testmini) | gemma-3-270m-it | unsloth | 81 | google/gemma-3-270m-it | 50 |
| gemma-3-270m-it | medium | [MathVista][] (testmini) | gemma-3-270m-it | unsloth | 82 | google/gemma-3-270m-it | 50 |
| gemma-3-270m-it-unsloth-bnb-4bit | medium | [MathVista][] (testmini) | gemma-3-270m-it | unsloth | 83 | google/gemma-3-270m-it | 50 |
| gemma-3-270m-it-qat-GGUF | medium | [MathVista][] (testmini) | gemma-3-270m-it-qat-q4-0-unquantized | unsloth | 84 | google/gemma-3-270m-it-qat-q4_0-unquantized | 50 |
| gemma-3-270m | medium | [MathVista][] (testmini) | gemma-3-270m | unsloth | 85 | google/gemma-3-270m | 50 |
| gemma-3-270m-unsloth-bnb-4bit | medium | [MathVista][] (testmini) | gemma-3-270m-it | unsloth | 86 | google/gemma-3-270m-it | 50 |
| gemma-3-270m-it-qat-q4_0-unquantized | medium | [MathVista][] (testmini) | gemma-3-270m-it | 87 | google/gemma-3-270m-it | 50 | |
| gemma-3-270m-it-bnb-4bit | medium | [MathVista][] (testmini) | gemma-3-270m-it | unsloth | 88 | google/gemma-3-270m-it | 50 |
| gemma-3-270m-it-qat | medium | [MathVista][] (testmini) | gemma-3-270m-it-qat-q4-0-unquantized | unsloth | 89 | google/gemma-3-270m-it-qat-q4_0-unquantized | 50 |
| gemma-3-270m-bnb-4bit | medium | [MathVista][] (testmini) | gemma-3-270m | unsloth | 90 | google/gemma-3-270m | 50 |
| gemma-3-270m-qat-q4_0-unquantized | medium | [MathVista][] (testmini) | gemma-3-270m | 91 | google/gemma-3-270m | 50 | |
| gemma-3-270m-it-GGUF | medium | [MathVista][] (testmini) | gemma-3-270m | Mungert | 92 | google/gemma-3-270m | 50 |
| gemma-3-270m-it-torchao-FP8 | medium | [MathVista][] (testmini) | gemma-3-270m-it | unsloth | 93 | google/gemma-3-270m-it | 50 |
| SmolVLM-Instruct-GGUF | high | Unknown | Unknown | Mungert | 94 | HuggingFaceTB/SmolLM2-1.7B-Instruct | 44.6 |
| Phi-3-vision-128k-instruct | medium | MathVista | Unknown | microsoft | 95 | microsoft/Phi-3-vision-128k-instruct | 44.5 |
| SmolVLM-256M-Instruct-GGUF | high | Unknown | Unknown | Mungert | 96 | HuggingFaceTB/SmolLM2-135M-Instruct | 35.9 |
Provenance & metadata (1)
source
- kind
- leaderboard-scrape
- paths
- benchmarks/mathvista.html