# GUI world modeling on MobileWorldBench Source: [GUI world modeling on MobileWorldBench]() Collection index: [llms.txt]() Data updated: 2026-05-12T15:32:53.518Z Evaluation of agent performance on GUI world modeling tasks using the MobileWorldBench dataset, measured by Accuracy (Acc.), Completeness (Com.), Relevance (Rele.), and Overall scores. 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 | GUI world modeling | | Dataset (including any stated split) | MobileWorldBench | | Metric | Accuracy | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 10 of 10 | | 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: SCALAR. - [Completeness](): Higher is better; unit: SCALAR. - [Relevance](): Higher is better; unit: SCALAR. - [Overall Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | MobileWorldModel-8B | 2.78 | Not recorded | 6a02b478dbc836b87d824bcf | | MobileWorldModel-8B | 2.66 | Not recorded | 6a02b478dbc836b87d824bcf | | GPT-5.2 | 2.62 | Not recorded | 6a02b478dbc836b87d824bcf | | Gemini-3.1-Pro | 2.57 | Not recorded | 6a02b478dbc836b87d824bcf | | MobileWorldModel-4B | 2.49 | Not recorded | 6a02b478dbc836b87d824bcf | | MobileWorldModel-4B | 2.47 | Not recorded | 6a02b478dbc836b87d824bcf | | Claude-4.6-Sonnet | 2.19 | Not recorded | 6a02b478dbc836b87d824bcf | | Qwen3-VL-4B | 1.98 | Not recorded | 6a02b478dbc836b87d824bcf | | Qwen3-VL-8B | 1.98 | Not recorded | 6a02b478dbc836b87d824bcf | | Qwen3-VL-32B | 1.93 | Not recorded | 6a02b478dbc836b87d824bcf | ### MobileWorldModel-8B Accuracy: 2.78. Result ID: 6a0347ce47f34ce924e87293. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"visual hint":"true","VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### MobileWorldModel-8B Accuracy: 2.66. Result ID: 6a0347ce47f34ce924e87292. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"visual hint":"false","VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### GPT-5.2 Accuracy: 2.62. Result ID: 6a0347ce47f34ce924e8728f. [Reporting paper]() Recorded evaluation configuration: {"VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### Gemini-3.1-Pro Accuracy: 2.57. Result ID: 6a0347ce47f34ce924e8728d. [Reporting paper]() Recorded evaluation configuration: {"VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### MobileWorldModel-4B Accuracy: 2.49. Result ID: 6a0347ce47f34ce924e87291. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"visual hint":"true","VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### MobileWorldModel-4B Accuracy: 2.47. Result ID: 6a0347ce47f34ce924e87290. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"visual hint":"false","VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### Claude-4.6-Sonnet Accuracy: 2.19. Result ID: 6a0347ce47f34ce924e8728e. [Reporting paper]() Recorded evaluation configuration: {"VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### Qwen3-VL-4B Accuracy: 1.98. Result ID: 6a0347ce47f34ce924e8728a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### Qwen3-VL-8B Accuracy: 1.98. Result ID: 6a0347ce47f34ce924e8728b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. ### Qwen3-VL-32B Accuracy: 1.93. Result ID: 6a0347ce47f34ce924e8728c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"VLM-as-a-judge":"Gemini-3.1-Pro"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.