# Image Understanding on MMEB Image v2 Source: [Image Understanding on MMEB Image v2]() Collection index: [llms.txt]() Data updated: 2026-03-04T16:57:01.326Z Evaluation of model performance on the MMEB Image v2 dataset focusing on general Image Understanding capabilities, including Classification (CLS) and Retrieval (RET) subtasks. 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 | Image Understanding | | Dataset (including any stated split) | MMEB Image v2 | | Metric | Accuracy (CLS) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 9 of 9 | | 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 (CLS)](): Higher is better; unit: PERCENT. - [Accuracy (RET)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (CLS) | Result date | Report paper | | --- | --- | --- | --- | | MetaEmbed | 68.1 | Not recorded | 69a66c572d603e6de3e3d146 | | B3++ | 67 | Not recorded | 69a66c572d603e6de3e3d146 | | VLM2VecV2 | 62.9 | Not recorded | 69a66c572d603e6de3e3d146 | | LLaVE | 62.1 | Not recorded | 69a66c572d603e6de3e3d146 | | UniME | 59 | Not recorded | 69a66c572d603e6de3e3d146 | | VLM2VecV1 | 58.7 | Not recorded | 69a66c572d603e6de3e3d146 | | GME | 54.4 | Not recorded | 69a66c572d603e6de3e3d146 | | OmniRet | 51.7 | Not recorded | 69a66c572d603e6de3e3d146 | | ColPaliv1.3 | 40.3 | Not recorded | 69a66c572d603e6de3e3d146 | ### MetaEmbed Accuracy (CLS): 68.1. Result ID: 69a8644e6dbb7075b9500ecf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"2B"}. This record does not assert independent reproduction. ### B3++ Accuracy (CLS): 67. Result ID: 69a8644e6dbb7075b9500ece. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"1.5B"}. This record does not assert independent reproduction. ### VLM2VecV2 Accuracy (CLS): 62.9. Result ID: 69a8644e6dbb7075b9500ed3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"1.5B"}. This record does not assert independent reproduction. ### LLaVE Accuracy (CLS): 62.1. Result ID: 69a8644e6dbb7075b9500ecd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"1.5B"}. This record does not assert independent reproduction. ### UniME Accuracy (CLS): 59. Result ID: 69a8644e6dbb7075b9500ecc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"1.5B"}. This record does not assert independent reproduction. ### VLM2VecV1 Accuracy (CLS): 58.7. Result ID: 69a8644e6dbb7075b9500ed2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"1.5B"}. This record does not assert independent reproduction. ### GME Accuracy (CLS): 54.4. Result ID: 69a8644e6dbb7075b9500ed1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"1.5B"}. This record does not assert independent reproduction. ### OmniRet Accuracy (CLS): 51.7. Result ID: 69a8644e6dbb7075b9500ed4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"1.5B"}. This record does not assert independent reproduction. ### ColPaliv1.3 Accuracy (CLS): 40.3. Result ID: 69a8644e6dbb7075b9500ed0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"LLM Size":"2B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.