# Multimodal Deep Search on MMSrch Source: [Multimodal Deep Search on MMSrch]() Collection index: [llms.txt]() Data updated: 2026-04-09T23:17:58.364Z Evaluation of model performance on the MMSrch dataset for Multimodal Deep Search tasks, 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 Deep Search | | Dataset (including any stated split) | MMSrch | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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 | | --- | --- | --- | --- | | Gemini-3 Pro | 82.94 | Not recorded | 69d732707d0172140ec76a0c | | MTA-DeepSearch-32B | 82.35 | Not recorded | 69d732707d0172140ec76a0c | | MTA-DeepSearch-8B | 79.41 | Not recorded | 69d732707d0172140ec76a0c | | GPT-5 | 77.65 | Not recorded | 69d732707d0172140ec76a0c | | Gemini-2.5 Pro | 77.65 | Not recorded | 69d732707d0172140ec76a0c | | SenseNova-MARS-32B | 74.27 | Not recorded | 69d732707d0172140ec76a0c | | MM-DeepResearch 32B | 69 | Not recorded | 69d732707d0172140ec76a0c | | Qwen3-VL-32B-Inst. | 68.52 | Not recorded | 69d732707d0172140ec76a0c | | SenseNova-MARS-8B | 67.84 | Not recorded | 69d732707d0172140ec76a0c | | Gemini-3-pro | 65.88 | Not recorded | 69d732707d0172140ec76a0c | | Qwen3-VL-8B-Inst. | 57.06 | Not recorded | 69d732707d0172140ec76a0c | | Webwatcher-32B | 55.3 | Not recorded | 69d732707d0172140ec76a0c | | MMSearch-R1-7B | 53.8 | Not recorded | 69d732707d0172140ec76a0c | | Webwatcher-7B | 49.1 | Not recorded | 69d732707d0172140ec76a0c | | Gemini-2.5-pro | 41.76 | Not recorded | 69d732707d0172140ec76a0c | | GPT-5 | 36.47 | Not recorded | 69d732707d0172140ec76a0c | | Qwen3-VL-32B-Inst. | 17.65 | Not recorded | 69d732707d0172140ec76a0c | | Qwen3-VL-8B-Inst. | 12.35 | Not recorded | 69d732707d0172140ec76a0c | ### Gemini-3 Pro Accuracy: 82.94. Result ID: 69d8334c47f34ce924bfdef3. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow"}. This record does not assert independent reproduction. ### MTA-DeepSearch-32B Accuracy: 82.35. Result ID: 69d8334c47f34ce924bfdefd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Ours","Model Scale":"32B"}. This record does not assert independent reproduction. ### MTA-DeepSearch-8B Accuracy: 79.41. Result ID: 69d8334c47f34ce924bfdefc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Ours","Model Scale":"8B"}. This record does not assert independent reproduction. ### GPT-5 Accuracy: 77.65. Result ID: 69d8334c47f34ce924bfdef1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow"}. This record does not assert independent reproduction. ### Gemini-2.5 Pro Accuracy: 77.65. Result ID: 69d8334c47f34ce924bfdef2. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow"}. This record does not assert independent reproduction. ### SenseNova-MARS-32B Accuracy: 74.27. Result ID: 69d8334c47f34ce924bfdef8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Protocol":"Multimodal Deep Search Agents","Model Scale":"32B"}. This record does not assert independent reproduction. ### MM-DeepResearch 32B Accuracy: 69. Result ID: 69d8334c47f34ce924bfdefb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Multimodal Deep Search Agents","Model Scale":"32B"}. This record does not assert independent reproduction. ### Qwen3-VL-32B-Inst. Accuracy: 68.52. Result ID: 69d8334c47f34ce924bfdef5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow","Model Scale":"32B"}. This record does not assert independent reproduction. ### SenseNova-MARS-8B Accuracy: 67.84. Result ID: 69d8334c47f34ce924bfdef7. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Protocol":"Multimodal Deep Search Agents","Model Scale":"8B"}. This record does not assert independent reproduction. ### Gemini-3-pro Accuracy: 65.88. Result ID: 69d8334c47f34ce924bfdeee. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Inst. Accuracy: 57.06. Result ID: 69d8334c47f34ce924bfdef4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow","Model Scale":"8B"}. This record does not assert independent reproduction. ### Webwatcher-32B Accuracy: 55.3. Result ID: 69d8334c47f34ce924bfdefa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Multimodal Deep Search Agents","Model Scale":"32B"}. This record does not assert independent reproduction. ### MMSearch-R1-7B Accuracy: 53.8. Result ID: 69d8334c47f34ce924bfdef6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Multimodal Deep Search Agents","Model Scale":"7B"}. This record does not assert independent reproduction. ### Webwatcher-7B Accuracy: 49.1. Result ID: 69d8334c47f34ce924bfdef9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Multimodal Deep Search Agents","Model Scale":"7B"}. This record does not assert independent reproduction. ### Gemini-2.5-pro Accuracy: 41.76. Result ID: 69d8334c47f34ce924bfdeed. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer"}. This record does not assert independent reproduction. ### GPT-5 Accuracy: 36.47. Result ID: 69d8334c47f34ce924bfdeec. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer"}. This record does not assert independent reproduction. ### Qwen3-VL-32B-Inst. Accuracy: 17.65. Result ID: 69d8334c47f34ce924bfdef0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer","Model Scale":"32B"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Inst. Accuracy: 12.35. Result ID: 69d8334c47f34ce924bfdeef. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer","Model Scale":"8B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.