# Multimodal Deep Search on FVQA (Accuracy) Source: [Multimodal Deep Search on FVQA (Accuracy)]() Collection index: [llms.txt]() Data updated: 2026-04-09T23:17:58.389Z Evaluation of model performance on the FVQA 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) | FVQA | | Metric | 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 - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Gemini-3 Pro | 76.67 | Not recorded | 69d732707d0172140ec76a0c | | MTA-DeepSearch-32B | 76 | Not recorded | 69d732707d0172140ec76a0c | | MTA-DeepSearch-8B | 73.06 | Not recorded | 69d732707d0172140ec76a0c | | SenseNova-MARS-32B | 72.61 | Not recorded | 69d732707d0172140ec76a0c | | Gemini-2.5 Pro | 72.33 | Not recorded | 69d732707d0172140ec76a0c | | GPT-5 | 72.28 | Not recorded | 69d732707d0172140ec76a0c | | MM-DeepResearch 32B | 70.1 | Not recorded | 69d732707d0172140ec76a0c | | SenseNova-MARS-8B | 67.11 | Not recorded | 69d732707d0172140ec76a0c | | Qwen3-VL-32B-Inst. | 66.94 | Not recorded | 69d732707d0172140ec76a0c | | Qwen3-VL-8B-Inst. | 64.22 | Not recorded | 69d732707d0172140ec76a0c | | Gemini-3-pro | 58.94 | Not recorded | 69d732707d0172140ec76a0c | | MMSearch-R1-7B | 58.4 | Not recorded | 69d732707d0172140ec76a0c | | Gemini-2.5-pro | 54.5 | Not recorded | 69d732707d0172140ec76a0c | | GPT-5 | 50.83 | Not recorded | 69d732707d0172140ec76a0c | | Qwen3-VL-8B-Inst. | 26.94 | Not recorded | 69d732707d0172140ec76a0c | | Qwen3-VL-32B-Inst. | 24.81 | Not recorded | 69d732707d0172140ec76a0c | ### Gemini-3 Pro Accuracy: 76.67. Result ID: 69d8334c47f34ce924bfdf05. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow"}. This record does not assert independent reproduction. ### MTA-DeepSearch-32B Accuracy: 76. Result ID: 69d8334c47f34ce924bfdf0d. [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: 73.06. Result ID: 69d8334c47f34ce924bfdf0c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Ours","Model Scale":"8B"}. This record does not assert independent reproduction. ### SenseNova-MARS-32B Accuracy: 72.61. Result ID: 69d8334c47f34ce924bfdf0a. [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. ### Gemini-2.5 Pro Accuracy: 72.33. Result ID: 69d8334c47f34ce924bfdf04. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow"}. This record does not assert independent reproduction. ### GPT-5 Accuracy: 72.28. Result ID: 69d8334c47f34ce924bfdf03. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow"}. This record does not assert independent reproduction. ### MM-DeepResearch 32B Accuracy: 70.1. Result ID: 69d8334c47f34ce924bfdf0b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Multimodal Deep Search Agents","Model Scale":"32B"}. This record does not assert independent reproduction. ### SenseNova-MARS-8B Accuracy: 67.11. Result ID: 69d8334c47f34ce924bfdf09. [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. ### Qwen3-VL-32B-Inst. Accuracy: 66.94. Result ID: 69d8334c47f34ce924bfdf07. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow","Model Scale":"32B"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Inst. Accuracy: 64.22. Result ID: 69d8334c47f34ce924bfdf06. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Agent Workflow","Model Scale":"8B"}. This record does not assert independent reproduction. ### Gemini-3-pro Accuracy: 58.94. Result ID: 69d8334c47f34ce924bfdf00. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer"}. This record does not assert independent reproduction. ### MMSearch-R1-7B Accuracy: 58.4. Result ID: 69d8334c47f34ce924bfdf08. [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: 54.5. Result ID: 69d8334c47f34ce924bfdeff. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer"}. This record does not assert independent reproduction. ### GPT-5 Accuracy: 50.83. Result ID: 69d8334c47f34ce924bfdefe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Inst. Accuracy: 26.94. Result ID: 69d8334c47f34ce924bfdf01. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer","Model Scale":"8B"}. This record does not assert independent reproduction. ### Qwen3-VL-32B-Inst. Accuracy: 24.81. Result ID: 69d8334c47f34ce924bfdf02. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Direct Answer","Model Scale":"32B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.