# Multimodal Reasoning and Perception on MMStar (test) Source: [Multimodal Reasoning and Perception on MMStar (test)]() Collection index: [llms.txt]() Data updated: 2026-04-07T10:32:08.030Z Evaluation of model performance on the test split of the MMStar dataset focusing on Multimodal Reasoning and Perception capabilities, measured by accuracy. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Multimodal Reasoning and Perception | | Dataset (including any stated split) | MMStar (test) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 19 of 19 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | GPT-4o | 63.9 | Not recorded | 69d48fb37d0172140ec744be | | InternVL2.5-8B | 63.2 | Not recorded | 69d48fb37d0172140ec744be | | Qwen2.5-VL-7B + Saliency-R1 | 62.6 | Not recorded | 69d48fb37d0172140ec744be | | Qwen2.5-VL-7B | 62.4 | Not recorded | 69d48fb37d0172140ec744be | | Claude-3.5 Sonnet | 62.2 | Not recorded | 69d48fb37d0172140ec744be | | LLaVA-OneVision-7B | 61.9 | Not recorded | 69d48fb37d0172140ec744be | | InternVL2-8B | 61.5 | Not recorded | 69d48fb37d0172140ec744be | | Insight-V-8B | 61.5 | Not recorded | 69d48fb37d0172140ec744be | | Vision-R1-LlamaV-CI-11B | 61.4 | Not recorded | 69d48fb37d0172140ec744be | | Mulberry-7B | 61.3 | Not recorded | 69d48fb37d0172140ec744be | | Qwen2.5-VL-7B + SFT | 60.1 | Not recorded | 69d48fb37d0172140ec744be | | IXC-2.5 | 59.9 | Not recorded | 69d48fb37d0172140ec744be | | Gemini-1.5-Pro | 59.1 | Not recorded | 69d48fb37d0172140ec744be | | MiniCPM-V-2.6-8B | 57.5 | Not recorded | 69d48fb37d0172140ec744be | | Qwen2.5-VL-3B | 56.1 | Not recorded | 69d48fb37d0172140ec744be | | Qwen2.5-VL-3B + Saliency-R1 | 56 | Not recorded | 69d48fb37d0172140ec744be | | Qwen2.5-VL-3B + SFT | 55.7 | Not recorded | 69d48fb37d0172140ec744be | | LLaVA-CoT-11B | 54.9 | Not recorded | 69d48fb37d0172140ec744be | | MiniCPM-Llama-V-2.5-8B | 51.8 | Not recorded | 69d48fb37d0172140ec744be | ### GPT-4o Accuracy: 63.9. Result ID: 69d4dd2447f34ce924bd54c3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### InternVL2.5-8B Accuracy: 63.2. Result ID: 69d4dd2447f34ce924bd54c7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2.5-VL-7B + Saliency-R1 Accuracy: 62.6. Result ID: 69d4dd2447f34ce924bd54d5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"Saliency-R1"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Accuracy: 62.4. Result ID: 69d4dd2447f34ce924bd54d3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"Base"}. This record does not assert independent reproduction. ### Claude-3.5 Sonnet Accuracy: 62.2. Result ID: 69d4dd2447f34ce924bd54c4. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### LLaVA-OneVision-7B Accuracy: 61.9. Result ID: 69d4dd2447f34ce924bd54c8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### InternVL2-8B Accuracy: 61.5. Result ID: 69d4dd2447f34ce924bd54c6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Insight-V-8B Accuracy: 61.5. Result ID: 69d4dd2447f34ce924bd54cd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Vision-R1-LlamaV-CI-11B Accuracy: 61.4. Result ID: 69d4dd2447f34ce924bd54cf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Mulberry-7B Accuracy: 61.3. Result ID: 69d4dd2447f34ce924bd54ce. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2.5-VL-7B + SFT Accuracy: 60.1. Result ID: 69d4dd2447f34ce924bd54d4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction. ### IXC-2.5 Accuracy: 59.9. Result ID: 69d4dd2447f34ce924bd54cb. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Gemini-1.5-Pro Accuracy: 59.1. Result ID: 69d4dd2447f34ce924bd54c5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MiniCPM-V-2.6-8B Accuracy: 57.5. Result ID: 69d4dd2447f34ce924bd54ca. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Qwen2.5-VL-3B Accuracy: 56.1. Result ID: 69d4dd2447f34ce924bd54d0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"Base"}. This record does not assert independent reproduction. ### Qwen2.5-VL-3B + Saliency-R1 Accuracy: 56. Result ID: 69d4dd2447f34ce924bd54d2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"Saliency-R1"}. This record does not assert independent reproduction. ### Qwen2.5-VL-3B + SFT Accuracy: 55.7. Result ID: 69d4dd2447f34ce924bd54d1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction. ### LLaVA-CoT-11B Accuracy: 54.9. Result ID: 69d4dd2447f34ce924bd54cc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MiniCPM-Llama-V-2.5-8B Accuracy: 51.8. Result ID: 69d4dd2447f34ce924bd54c9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.