# GUI Navigation on Multimodal-Mind2Web Cross-Domain Source: [GUI Navigation on Multimodal-Mind2Web Cross-Domain]() Collection index: [llms.txt]() Data updated: 2026-04-23T08:28:14.978Z Evaluation of GUI navigation performance on the Multimodal-Mind2Web Cross-Domain dataset, reporting metrics including Element Accuracy (Ele.Acc), Operation F1 (Op.F1), and Step Success Rate (Step SR). 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 Navigation | | Dataset (including any stated split) | Multimodal-Mind2Web Cross-Domain | | Metric | Step Success Rate | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 32 of 32 | | 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 - [Step Success Rate](): Higher is better; unit: PERCENT. - [Element Accuracy](): Higher is better; unit: PERCENT. - [Operation F1](): Higher is better; unit: PERCENT. ## Recorded results | Method | Step Success Rate | Result date | Report paper | | --- | --- | --- | --- | | HTML-T5-XL | 67.1 | Not recorded | 68fa78041e209131cea765a6 | | CogAgent | 59.4 | Not recorded | 68fa78041e209131cea765a6 | | AutoWebGLM | 55.8 | Not recorded | 68fa78041e209131cea765a6 | | Explorer-7B | 53 | Not recorded | 68fa78e41e209131cea7b049 | | AgentTrek-7B | 52.6 | Not recorded | 68fa78e41e209131cea7b049 | | LLaMA2-7B | 50.3 | Not recorded | 68fa78041e209131cea765a6 | | LPO | 49.6 | Not recorded | 68fa79731e209131cea7db23 | | SpiritSight-26B | 49.2 | Not recorded | 68fa78041e209131cea765a6 | | Explorer-7B | 48.1 | Not recorded | 68fa78e41e209131cea7b049 | | RGUI-R1 | 47.9 | Not recorded | 68fa79731e209131cea7db23 | | Explorer-4B | 47.2 | Not recorded | 68fa78e41e209131cea7b049 | | Explorer-7B | 44.7 | Not recorded | 68fa78e41e209131cea7b049 | | Explorer-4B | 44.6 | Not recorded | 68fa78e41e209131cea7b049 | | SpiritSight-8B | 44.4 | Not recorded | 68fa78041e209131cea765a6 | | OmniParser | 42 | Not recorded | 68fa78041e209131cea765a6 | | Base Model | 40.7 | Not recorded | 68fa79731e209131cea7db23 | | RInfiGUI-R1 | 40 | Not recorded | 68fa79731e209131cea7db23 | | Explorer-4B | 39.8 | Not recorded | 68fa78e41e209131cea7b049 | | ScribeAgent-32B | 37.3 | Not recorded | 68fa78e41e209131cea7b049 | | SpiritSight-2B | 36.9 | Not recorded | 68fa78041e209131cea765a6 | | SeeAct | 36.8 | Not recorded | 68fa78041e209131cea765a6 | | SeeAct | 36.8 | Not recorded | 68fa78e41e209131cea7b049 | | ReadAgent-P | 33.4 | Not recorded | 68fa78041e209131cea765a6 | | GPT-4 | 29.7 | Not recorded | 68fa78e41e209131cea7b049 | | RUI-R1 | 27.1 | Not recorded | 68fa79731e209131cea7db23 | | GPT-3.5 | 24.1 | Not recorded | 68fa78e41e209131cea7b049 | | EDGE-9.6B | 22.4 | Not recorded | 68fa78e41e209131cea7b049 | | SeeClick | 20.8 | Not recorded | 68fa78041e209131cea765a6 | | SeeClick-9.6B | 20.2 | Not recorded | 68fa78e41e209131cea7b049 | | MiniCPM-GUI | 14.6 | Not recorded | 68fa78041e209131cea765a6 | | MiniCPM-3.1B | 14.6 | Not recorded | 68fa78e41e209131cea7b049 | | Fuyu-GUI | 11.7 | Not recorded | 68fa78041e209131cea765a6 | ### HTML-T5-XL Step Success Rate: 67.1. Result ID: 6995786c3badfaffe66f0cf1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B","Input Modality":"Text","Select From Top":true}. This record does not assert independent reproduction. ### CogAgent Step Success Rate: 59.4. Result ID: 6995786c3badfaffe66f0cf0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"18B","Input Modality":"Image","Select From Top":true}. This record does not assert independent reproduction. ### AutoWebGLM Step Success Rate: 55.8. Result ID: 6995786c3badfaffe66f0cee. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"6B","Input Modality":"Text","Select From Top":true}. This record does not assert independent reproduction. ### Explorer-7B Step Success Rate: 53. Result ID: 6995ab2c3badfaffe6a9aa1f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. + M2W"}. This record does not assert independent reproduction. ### AgentTrek-7B Step Success Rate: 52.6. Result ID: 6995ab2c3badfaffe6a9aa19. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. + M2W"}. This record does not assert independent reproduction. ### LLaMA2-7B Step Success Rate: 50.3. Result ID: 6995786c3badfaffe66f0cef. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B","Input Modality":"Text","Select From Top":true}. This record does not assert independent reproduction. ### LPO Step Success Rate: 49.6. Result ID: 69e9d81647f34ce924d0f8c5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"After Preference Optimization"}. This record does not assert independent reproduction. ### SpiritSight-26B Step Success Rate: 49.2. Result ID: 6995786c3badfaffe66f0cfa. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"26B","Input Modality":"Image","Select From Top":false}. This record does not assert independent reproduction. ### Explorer-7B Step Success Rate: 48.1. Result ID: 6995ab2c3badfaffe6a9aa1e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"M2W"}. This record does not assert independent reproduction. ### RGUI-R1 Step Success Rate: 47.9. Result ID: 69e9d81647f34ce924d0f8c3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"After Preference Optimization"}. This record does not assert independent reproduction. ### Explorer-4B Step Success Rate: 47.2. Result ID: 6995ab2c3badfaffe6a9aa1c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. + M2W"}. This record does not assert independent reproduction. ### Explorer-7B Step Success Rate: 44.7. Result ID: 6995ab2c3badfaffe6a9aa1d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. traj."}. This record does not assert independent reproduction. ### Explorer-4B Step Success Rate: 44.6. Result ID: 6995ab2c3badfaffe6a9aa1b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"M2W"}. This record does not assert independent reproduction. ### SpiritSight-8B Step Success Rate: 44.4. Result ID: 6995786c3badfaffe66f0cf9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"8B","Input Modality":"Image","Select From Top":false}. This record does not assert independent reproduction. ### OmniParser Step Success Rate: 42. Result ID: 6995786c3badfaffe66f0cf7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Input Modality":"Image","Select From Top":false}. This record does not assert independent reproduction. ### Base Model Step Success Rate: 40.7. Result ID: 69e9d81647f34ce924d0f8c1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"After Supervised Fine-Tuning"}. This record does not assert independent reproduction. ### RInfiGUI-R1 Step Success Rate: 40. Result ID: 69e9d81647f34ce924d0f8c4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"After Preference Optimization"}. This record does not assert independent reproduction. ### Explorer-4B Step Success Rate: 39.8. Result ID: 6995ab2c3badfaffe6a9aa1a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. traj."}. This record does not assert independent reproduction. ### ScribeAgent-32B Step Success Rate: 37.3. Result ID: 6995ab2c3badfaffe6a9aa18. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. traj."}. This record does not assert independent reproduction. ### SpiritSight-2B Step Success Rate: 36.9. Result ID: 6995786c3badfaffe66f0cf8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"2B","Input Modality":"Image","Select From Top":false}. This record does not assert independent reproduction. ### SeeAct Step Success Rate: 36.8. Result ID: 6995786c3badfaffe66f0cf2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Input Modality":"Text+Image","Select From Top":false}. This record does not assert independent reproduction. ### SeeAct Step Success Rate: 36.8. Result ID: 6995ab2c3badfaffe6a9aa14. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"In-Context Learning"}. This record does not assert independent reproduction. ### ReadAgent-P Step Success Rate: 33.4. Result ID: 6995786c3badfaffe66f0cf3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"340B","Input Modality":"Text","Select From Top":false}. This record does not assert independent reproduction. ### GPT-4 Step Success Rate: 29.7. Result ID: 6995ab2c3badfaffe6a9aa13. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Protocol":"In-Context Learning"}. This record does not assert independent reproduction. ### RUI-R1 Step Success Rate: 27.1. Result ID: 69e9d81647f34ce924d0f8c2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Protocol":"After Preference Optimization"}. This record does not assert independent reproduction. ### GPT-3.5 Step Success Rate: 24.1. Result ID: 6995ab2c3badfaffe6a9aa12. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"In-Context Learning"}. This record does not assert independent reproduction. ### EDGE-9.6B Step Success Rate: 22.4. Result ID: 6995ab2c3badfaffe6a9aa16. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. + M2W"}. This record does not assert independent reproduction. ### SeeClick Step Success Rate: 20.8. Result ID: 6995786c3badfaffe66f0cf6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"9.6B","Input Modality":"Image","Select From Top":false}. This record does not assert independent reproduction. ### SeeClick-9.6B Step Success Rate: 20.2. Result ID: 6995ab2c3badfaffe6a9aa15. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. + M2W"}. This record does not assert independent reproduction. ### MiniCPM-GUI Step Success Rate: 14.6. Result ID: 6995786c3badfaffe66f0cf4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Size":"3B","Input Modality":"Image","Select From Top":false}. This record does not assert independent reproduction. ### MiniCPM-3.1B Step Success Rate: 14.6. Result ID: 6995ab2c3badfaffe6a9aa17. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Protocol":"Supervised Fine-Tuning","Train Data":"Syn. + M2W"}. This record does not assert independent reproduction. ### Fuyu-GUI Step Success Rate: 11.7. Result ID: 6995786c3badfaffe66f0cf5. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"8B","Input Modality":"Image","Select From Top":false}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.