# Multimodal Reasoning on MMstar (Pass@1 accuracy) Source: [Multimodal Reasoning on MMstar (Pass@1 accuracy)]() Collection index: [llms.txt]() Data updated: 2026-04-14T08:50:59.837Z Evaluation of model performance on the MMStar dataset for Multimodal Reasoning, specifically reporting Pass@1 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 | | Dataset (including any stated split) | MMstar | | Metric | Pass@1 Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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 - [Pass@1 Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Pass@1 Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Qwen2.5-VL-7B-Instruct + RFT | 67.1 | Not recorded | 69ddca347d0172140ec79b4e | | InternVL2.5-26B | 66.5 | Not recorded | 69ddca347d0172140ec79b4e | | InternVL3-9B | 66.3 | Not recorded | 69ddca347d0172140ec79b4e | | Qwen2.5-VL-7B-Instruct + cold start | 66.2 | Not recorded | 69ddca347d0172140ec79b4e | | LLaVA-OneVision-72B | 65.8 | Not recorded | 69ddca347d0172140ec79b4e | | Claude-3.5-Sonnet | 65.1 | Not recorded | 69ddca347d0172140ec79b4e | | GPT-4o-20240513 | 64.7 | Not recorded | 69ddca347d0172140ec79b4e | | Qwen2.5-VL-7B-Instruct | 63.9 | Not recorded | 69ddca347d0172140ec79b4e | | Qwen2.5-VL-7B-Instruct | 63.9 | Not recorded | 69ddca347d0172140ec79b4e | | InternVL2.5-8B | 62.8 | Not recorded | 69ddca347d0172140ec79b4e | | Qwen2.5-VL-7B-Instruct | 61.8 | Not recorded | 69ddca347d0172140ec79b4e | | Qwen2-VL-7B | 60.7 | Not recorded | 69ddca347d0172140ec79b4e | | Gemini-1.5-Pro | 59.1 | Not recorded | 69ddca347d0172140ec79b4e | | MiniCPM-V2.6 | 57.5 | Not recorded | 69ddca347d0172140ec79b4e | | GPT-4V | 56 | Not recorded | 69ddca347d0172140ec79b4e | | Cambrian-34B | 54.2 | Not recorded | 69ddca347d0172140ec79b4e | ### Qwen2.5-VL-7B-Instruct + RFT Pass@1 Accuracy: 67.1. Result ID: 69ddffa047f34ce924c582ac. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Ours","Training Strategy":"RFT"}. This record does not assert independent reproduction. ### InternVL2.5-26B Pass@1 Accuracy: 66.5. Result ID: 69ddffa047f34ce924c582a7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Source Category":"Open-Source Models"}. This record does not assert independent reproduction. ### InternVL3-9B Pass@1 Accuracy: 66.3. Result ID: 69ddffa047f34ce924c582a3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Open-Source Models"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B-Instruct + cold start Pass@1 Accuracy: 66.2. Result ID: 69ddffa047f34ce924c582ab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Ours","Training Strategy":"cold start"}. This record does not assert independent reproduction. ### LLaVA-OneVision-72B Pass@1 Accuracy: 65.8. Result ID: 69ddffa047f34ce924c582a6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Source Category":"Open-Source Models"}. This record does not assert independent reproduction. ### Claude-3.5-Sonnet Pass@1 Accuracy: 65.1. Result ID: 69ddffa047f34ce924c5829f. [Reporting paper]() Recorded evaluation configuration: {"Model Source Category":"Closed-Source Models"}. This record does not assert independent reproduction. ### GPT-4o-20240513 Pass@1 Accuracy: 64.7. Result ID: 69ddffa047f34ce924c5829e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Closed-Source Models"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B-Instruct Pass@1 Accuracy: 63.9. Result ID: 69ddffa047f34ce924c582a8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Base Model"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B-Instruct Pass@1 Accuracy: 63.9. Result ID: 69ddffa047f34ce924c582aa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Ours"}. This record does not assert independent reproduction. ### InternVL2.5-8B Pass@1 Accuracy: 62.8. Result ID: 69ddffa047f34ce924c582a1. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Source Category":"Open-Source Models"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B-Instruct Pass@1 Accuracy: 61.8. Result ID: 69ddffa047f34ce924c582a9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Base Model","Chain-of-Thought (CoT)":"true"}. This record does not assert independent reproduction. ### Qwen2-VL-7B Pass@1 Accuracy: 60.7. Result ID: 69ddffa047f34ce924c582a4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Source Category":"Open-Source Models"}. This record does not assert independent reproduction. ### Gemini-1.5-Pro Pass@1 Accuracy: 59.1. Result ID: 69ddffa047f34ce924c582a0. [Reporting paper]() Recorded evaluation configuration: {"Model Source Category":"Closed-Source Models"}. This record does not assert independent reproduction. ### MiniCPM-V2.6 Pass@1 Accuracy: 57.5. Result ID: 69ddffa047f34ce924c582a2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Open-Source Models"}. This record does not assert independent reproduction. ### GPT-4V Pass@1 Accuracy: 56. Result ID: 69ddffa047f34ce924c5829d. [Reporting paper]() Recorded evaluation configuration: {"Model Source Category":"Closed-Source Models"}. This record does not assert independent reproduction. ### Cambrian-34B Pass@1 Accuracy: 54.2. Result ID: 69ddffa047f34ce924c582a5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Source Category":"Open-Source Models"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.