# GUI Grounding on OSWorld-G

Source: [GUI Grounding on OSWorld-G](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-07-01T07:55:53.048Z

Evaluation of GUI Grounding performance on the OSWorld-G dataset, reporting component-level metrics like Text Matching, Element Recognition, Layout Understanding, and Fine-grained Manipulation.

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 Grounding |
| Dataset (including any stated split) | OSWorld-G |
| Metric | Average Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 267 of 267 |
| 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

- [Average Score](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

- [Text Matching](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Text%20Matching>): Higher is better; unit: PERCENT.

- [Element Recognition](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Element%20Recognition>): Higher is better; unit: PERCENT.

- [Layout Understanding](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Layout%20Understanding>): Higher is better; unit: PERCENT.

- [Fine-grained Manipulation](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Fine-grained%20Manipulation>): Higher is better; unit: PERCENT.

- [Delta/Change](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Delta%2FChange>): Higher is better; unit: SCALAR.

- [Refusal Rate](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Refusal%20Rate>): Lower is better; unit: PERCENT.

- [Accuracy](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Top-1 Accuracy](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Top-1%20Accuracy>): Higher is better; unit: PERCENT.

- [Latency](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Latency>): Lower is better; unit: TIME.

## Recorded results

| Method | Average Score | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen3VL-8B-Instruct | 72.7 | Not recorded | 6938fe876cfd488761a53c83 |
| Qwen3VL-32B-Instruct | 72 | Not recorded | 6938fe876cfd488761a53c83 |
| Qwen3VL-32B-Instruct | 71.7 | Not recorded | 6938fe876cfd488761a53c83 |
| UI-Venus-1.5-30B-A3B | 70.6 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| UI-Venus-1.5-30B-A3B | 70.6 | Not recorded | 6a449f10e0224bd4f4993095 |
| UI-Venus-Ground-72B | 70.4 | Not recorded | 6938fe876cfd488761a53c83 |
| Qwen3-VL-32B-Instruct + Re-Prefill | 70.1 | Not recorded | 6a05570edbc836b87d828c27 |
| Step-GUI-8B | 70 | Not recorded | 69438a9b423ea422657294f6 |
| UI-Venus-1.5-8B | 69.7 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| UI-Venus-1.5-8B | 69.7 | Not recorded | 6a449f10e0224bd4f4993095 |
| Qwen3-VL-32B-Instruct | 69 | Not recorded | 6a05570edbc836b87d828c27 |
| Qwen3VL-8B-Instruct | 68.8 | Not recorded | 6938fe876cfd488761a53c83 |
| GUI-Owl-1.5-8B-Instruct + Re-Prefill | 68.8 | Not recorded | 6a05570edbc836b87d828c27 |
| GTA1-7B | 68.7 | Not recorded | 6938fe876cfd488761a53c83 |
| MAI-UI-8B + Re-Prefill | 67.7 | Not recorded | 6a05570edbc836b87d828c27 |
| GTA1-7B | 67.7 | Not recorded | 68fa7b911e209131cea892ed |
| EvoCUA-OpenCUA-72B | 67.65 | Not recorded | 6973007bb2be0ed85a78c9c4 |
| MAI-UI-32B | 67.6 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| Mimo-VL-7B + WildGUI | 67.6 | Not recorded | 6a06a8b7dbc836b87d82a1de |
| MAI-UI-32B | 67.6 | Not recorded | 6a449f10e0224bd4f4993095 |
| GTA1-7B | 67.5 | Not recorded | 6938fe876cfd488761a53c83 |
| GUI-Owl-1.5-8B-Instruct | 67.2 | Not recorded | 6a05570edbc836b87d828c27 |
| OpenCUA-72B | 66.95 | Not recorded | 6973007bb2be0ed85a78c9c4 |
| Step-GUI-4B | 66.9 | Not recorded | 69438a9b423ea422657294f6 |
| UI-TARS-1.5-7B | 66.8 | Not recorded | 6938fe876cfd488761a53c83 |
| GUI-Owl-1.5-32B-Instruct | 66.8 | Not recorded | 6a449f10e0224bd4f4993095 |
| GTA1-72B | 66.7 | Not recorded | 6938fe876cfd488761a53c83 |
| GUI-Owl-1.5-8B-Instruct | 65.8 | Not recorded | 6a05570edbc836b87d828c27 |
| GUI-Owl-1.5-8B-Instruct | 65.8 | Not recorded | 6a449f10e0224bd4f4993095 |
| Qwen3-VL-8B-Instruct + Re-Prefill | 65.7 | Not recorded | 6a05570edbc836b87d828c27 |
| GUI-Cursor | 65.6 | Not recorded | 68fa7b911e209131cea892ed |
| MAI-UI-8B | 65.5 | Not recorded | 6a05570edbc836b87d828c27 |
| GTA1-32B | 65.2 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| Holo2-30B-A3B | 65.2 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| GTA1-32B | 65.2 | Not recorded | 6a05570edbc836b87d828c27 |
| MAI-UI-8B | 64.2 | Not recorded | 6a05570edbc836b87d828c27 |
| Qwen3-VL-32B-Thinking | 64 | Not recorded | 6973007bb2be0ed85a78c9c4 |
| EvoCUA-32B | 63.86 | Not recorded | 6973007bb2be0ed85a78c9c4 |
| GUI-Owl-1.5-4B-Instruct | 63.7 | Not recorded | 6a449f10e0224bd4f4993095 |
| Holo2-8B | 63.5 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| Qwen3-VL-8B-Instruct | 63 | Not recorded | 6a05570edbc836b87d828c27 |
| Seed1.5-VL | 62.9 | Not recorded | 6938fe876cfd488761a53c83 |
| Seed1.5-VL | 62.9 | Not recorded | 6a06a8b7dbc836b87d82a1de |
| GUI-Spotlight | 62.7 | Not recorded | 6938fe876cfd488761a53c83 |
| GUI-Spotlight | 62.7 | Not recorded | 68fa7b911e209131cea892ed |
| UI-Venus-1.0-72B | 62.2 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| Qwen2.5-VL-72B | 62.2 | Not recorded | 6938fe876cfd488761a53c83 |
| UI-Venus-Ground-72B | 62.2 | Not recorded | 6a449f10e0224bd4f4993095 |
| UI-TARS-1.5-7B | 61.9 | Not recorded | 6938fe876cfd488761a53c83 |
| UI-TARS-1.5-7B | 61.9 | Not recorded | 68fa7b911e209131cea892ed |
| Qwen3-VL-30B-A3B | 61.2 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| Qwen3-VL-30B-A3B | 61.2 | Not recorded | 6a449f10e0224bd4f4993095 |
| Qwen3-30B-A3B-Instruct | 61 | Not recorded | 69438a9b423ea422657294f6 |
| Qwen3-VL-32B | 60.6 | Not recorded | 6a06a8b7dbc836b87d82a1de |
| Step-GUI-4B | 60.5 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| Step-GUI-4B | 60.5 | Not recorded | 6a449f10e0224bd4f4993095 |
| GTA1-7B | 60.1 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| MAI-UI-8B | 60.1 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| MAI-UI-8B | 60.1 | Not recorded | 6a449f10e0224bd4f4993095 |
| OpenCUA-32B | 59.6 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| UI-Venus-1.5-2B | 59.4 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| UI-Venus-7B | 58.8 | Not recorded | 6a06a8b7dbc836b87d82a1de |
| Xiaomi-GUI-0-30B-A3B | 58.7 | Not recorded | 6a449f10e0224bd4f4993095 |
| Trifuse+GUI-AIMA | 58.4 | Not recorded | 698969d7c97a0b3e3b5dc31f |
| GUI-AIMA | 58.3 | Not recorded | 698969d7c97a0b3e3b5dc31f |
| Qwen3-VL-4B-Instruct | 58.2 | Not recorded | 69438a9b423ea422657294f6 |
| Qwen3-VL-8B-Instruct | 58.2 | Not recorded | 69438a9b423ea422657294f6 |
| GUI-Owl-32B | 58 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| GUI-Owl-32B | 58 | Not recorded | 69438a9b423ea422657294f6 |
| GUI-Cursor | 58 | Not recorded | 68fa7b911e209131cea892ed |
| GUI-Owl-32B | 58 | Not recorded | 6a449f10e0224bd4f4993095 |
| Trifuse+GUI-Actor | 57.9 | Not recorded | 698969d7c97a0b3e3b5dc31f |
| GUI-Owl-1.5-32B-Thinking | 57.6 | Not recorded | 6a449f10e0224bd4f4993095 |
| Qwen3-VL-8B | 57.5 | Not recorded | 698c0ceec97a0b3e3b5e104f |
| Qwen3-VL-8B | 57.5 | Not recorded | 6a449f10e0224bd4f4993095 |
| UI-TARS-72B | 57.1 | Not recorded | 6938fe876cfd488761a53c83 |
| UI-TARS-72B | 57.1 | Not recorded | 6a05570edbc836b87d828c27 |
| UI-TARS-72B | 57.1 | Not recorded | 6a06a8b7dbc836b87d82a1de |
| UI-TARS-72B | 57.1 | Not recorded | 6a449f10e0224bd4f4993095 |
| Qwen3-VL-8B-Thinking | 56.7 | Not recorded | 6973007bb2be0ed85a78c9c4 |

### Qwen3VL-8B-Instruct

Average Score: 72.7. Result ID: 69986e745698665ec4ce5c92.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

Recorded evaluation configuration: {"Access":"Open-Source","MVP Framework":true}. This record does not assert independent reproduction.

### Qwen3VL-32B-Instruct

Average Score: 72. Result ID: 69986e745698665ec4ce5c94.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

Recorded evaluation configuration: {"Access":"Open-Source","MVP Framework":true}. This record does not assert independent reproduction.

### Qwen3VL-32B-Instruct

Average Score: 71.7. Result ID: 69986e745698665ec4ce5c93.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

Recorded evaluation configuration: {"Access":"Open-Source","MVP Framework":false}. This record does not assert independent reproduction.

### UI-Venus-1.5-30B-A3B

Average Score: 70.6. Result ID: 699722185698665ec4705c12.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Implementation](<https://github.com/inclusionAI/UI-Venus>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### UI-Venus-1.5-30B-A3B

Average Score: 70.6. Result ID: 6a44c80873d097d845bedef0.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Implementation](<https://github.com/inclusionAI/UI-Venus>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"30B-A3B","Version":"1.5"}. This record does not assert independent reproduction.

### UI-Venus-Ground-72B

Average Score: 70.4. Result ID: 69986e745698665ec4ce5c86.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ab01e209131cea849f9.md>)

Recorded evaluation configuration: {"Access":"Open-Source"}. This record does not assert independent reproduction.

### Qwen3-VL-32B-Instruct + Re-Prefill

Average Score: 70.1. Result ID: 6a0577c447f34ce924ec0f87.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: {"zoom-in strategy":"true","Re-Prefill":"true"}. This record does not assert independent reproduction.

### Step-GUI-8B

Average Score: 70. Result ID: 699837d15698665ec4938e87.

[Reporting paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Implementation](<https://github.com/stepfun-ai/gelab-zero>)

Recorded evaluation configuration: {"Parameters":"8B"}. This record does not assert independent reproduction.

### UI-Venus-1.5-8B

Average Score: 69.7. Result ID: 699722185698665ec4705c11.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Implementation](<https://github.com/inclusionAI/UI-Venus>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### UI-Venus-1.5-8B

Average Score: 69.7. Result ID: 6a44c80873d097d845bedeef.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Implementation](<https://github.com/inclusionAI/UI-Venus>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"8B","Version":"1.5"}. This record does not assert independent reproduction.

### Qwen3-VL-32B-Instruct

Average Score: 69. Result ID: 6a0577c447f34ce924ec0f86.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: {"zoom-in strategy":"true"}. This record does not assert independent reproduction.

### Qwen3VL-8B-Instruct

Average Score: 68.8. Result ID: 69986e745698665ec4ce5c91.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

Recorded evaluation configuration: {"Access":"Open-Source","MVP Framework":false}. This record does not assert independent reproduction.

### GUI-Owl-1.5-8B-Instruct + Re-Prefill

Average Score: 68.8. Result ID: 6a0577c447f34ce924ec0f85.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: {"zoom-in strategy":"true","Re-Prefill":"true"}. This record does not assert independent reproduction.

### GTA1-7B

Average Score: 68.7. Result ID: 69986e745698665ec4ce5c90.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

Recorded evaluation configuration: {"Access":"Open-Source","MVP Framework":true}. This record does not assert independent reproduction.

### MAI-UI-8B + Re-Prefill

Average Score: 67.7. Result ID: 6a0577c447f34ce924ec0f83.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: {"zoom-in strategy":"true","Re-Prefill":"true"}. This record does not assert independent reproduction.

### GTA1-7B

Average Score: 67.7. Result ID: 6a169e16fbc373592c0f6bd8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b911e209131cea892ed.md>)

Recorded evaluation configuration: {"Initialised from":"UI-TARS-1.5-7B"}. This record does not assert independent reproduction.

### EvoCUA-OpenCUA-72B

Average Score: 67.65. Result ID: 699623e45698665ec464825c.

[Reporting paper](<https://www.sota2.com/research/paper/6973007bb2be0ed85a78c9c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6973007bb2be0ed85a78c9c4.md>)

[Implementation](<https://github.com/meituan/EvoCUA>)

Recorded evaluation configuration: {"Base Model":"OpenCUA-72B"}. This record does not assert independent reproduction.

### MAI-UI-32B

Average Score: 67.6. Result ID: 699722185698665ec4705c0f.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69520ae053fdd5c36173d429.md>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### Mimo-VL-7B + WildGUI

Average Score: 67.6. Result ID: 6a06e3c947f34ce924eeca4d.

[Reporting paper](<https://www.sota2.com/research/paper/6a06a8b7dbc836b87d82a1de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a06a8b7dbc836b87d82a1de.md>)

[Implementation](<https://github.com/WeiminXiong/Video2GUI>)

Recorded evaluation configuration: {"Model Category":"Effectiveness of WildGUI (Ours)","Training":"Continual Pre-training on WildGUI"}. This record does not assert independent reproduction.

### MAI-UI-32B

Average Score: 67.6. Result ID: 6a44c80873d097d845bedefa.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69520ae053fdd5c36173d429.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"32B"}. This record does not assert independent reproduction.

### GTA1-7B

Average Score: 67.5. Result ID: 69986e745698665ec4ce5c8f.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

Recorded evaluation configuration: {"Access":"Open-Source","MVP Framework":false}. This record does not assert independent reproduction.

### GUI-Owl-1.5-8B-Instruct

Average Score: 67.2. Result ID: 6a0577c447f34ce924ec0f84.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: {"zoom-in strategy":"true"}. This record does not assert independent reproduction.

### OpenCUA-72B

Average Score: 66.95. Result ID: 699623e45698665ec4648259.

[Reporting paper](<https://www.sota2.com/research/paper/6973007bb2be0ed85a78c9c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b2f4.md>)

Recorded evaluation configuration: {"Parameters":"72B"}. This record does not assert independent reproduction.

### Step-GUI-4B

Average Score: 66.9. Result ID: 699837d15698665ec4938e86.

[Reporting paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Implementation](<https://github.com/stepfun-ai/gelab-zero>)

Recorded evaluation configuration: {"Parameters":"4B"}. This record does not assert independent reproduction.

### UI-TARS-1.5-7B

Average Score: 66.8. Result ID: 69986e745698665ec4ce5c8e.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

Recorded evaluation configuration: {"Access":"Open-Source","MVP Framework":true}. This record does not assert independent reproduction.

### GUI-Owl-1.5-32B-Instruct

Average Score: 66.8. Result ID: 6a44c80873d097d845bedef6.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6997eaac09c33900fcd766ce.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"32B","Version":"1.5","Type":"Instruct"}. This record does not assert independent reproduction.

### GTA1-72B

Average Score: 66.7. Result ID: 69986e745698665ec4ce5c8b.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b2ba.md>)

Recorded evaluation configuration: {"Access":"Open-Source"}. This record does not assert independent reproduction.

### GUI-Owl-1.5-8B-Instruct

Average Score: 65.8. Result ID: 6a0577c447f34ce924ec0f7f.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### GUI-Owl-1.5-8B-Instruct

Average Score: 65.8. Result ID: 6a44c80873d097d845bedef4.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6997eaac09c33900fcd766ce.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"8B","Version":"1.5","Type":"Instruct"}. This record does not assert independent reproduction.

### Qwen3-VL-8B-Instruct + Re-Prefill

Average Score: 65.7. Result ID: 6a0577c447f34ce924ec0f81.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: {"zoom-in strategy":"true","Re-Prefill":"true"}. This record does not assert independent reproduction.

### GUI-Cursor

Average Score: 65.6. Result ID: 6a169e16fbc373592c0f6bda.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b911e209131cea892ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b911e209131cea892ed.md>)

Recorded evaluation configuration: {"Initialised from":"UI-TARS-1.5-7B"}. This record does not assert independent reproduction.

### MAI-UI-8B

Average Score: 65.5. Result ID: 6a0577c447f34ce924ec0f82.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: {"zoom-in strategy":"true"}. This record does not assert independent reproduction.

### GTA1-32B

Average Score: 65.2. Result ID: 699722185698665ec4705c04.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b2ba.md>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### Holo2-30B-A3B

Average Score: 65.2. Result ID: 699722185698665ec4705c0b.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

Recorded evaluation configuration: {"reproduction":true,"design":"end-to-end"}. This record does not assert independent reproduction.

### GTA1-32B

Average Score: 65.2. Result ID: 6a0577c447f34ce924ec0f7d.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b2ba.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### MAI-UI-8B

Average Score: 64.2. Result ID: 6a0577c447f34ce924ec0f7e.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Qwen3-VL-32B-Thinking

Average Score: 64. Result ID: 699623e45698665ec464825b.

[Reporting paper](<https://www.sota2.com/research/paper/6973007bb2be0ed85a78c9c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Parameters":"32B","Thinking Mode":true}. This record does not assert independent reproduction.

### EvoCUA-32B

Average Score: 63.86. Result ID: 699623e45698665ec464825e.

[Reporting paper](<https://www.sota2.com/research/paper/6973007bb2be0ed85a78c9c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6973007bb2be0ed85a78c9c4.md>)

[Implementation](<https://github.com/meituan/EvoCUA>)

Recorded evaluation configuration: {"Parameters":"32B"}. This record does not assert independent reproduction.

### GUI-Owl-1.5-4B-Instruct

Average Score: 63.7. Result ID: 6a44c80873d097d845bedef3.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6997eaac09c33900fcd766ce.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"4B","Version":"1.5","Type":"Instruct"}. This record does not assert independent reproduction.

### Holo2-8B

Average Score: 63.5. Result ID: 699722185698665ec4705c0a.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

Recorded evaluation configuration: {"reproduction":true,"design":"end-to-end"}. This record does not assert independent reproduction.

### Qwen3-VL-8B-Instruct

Average Score: 63. Result ID: 6a0577c447f34ce924ec0f80.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

Recorded evaluation configuration: {"zoom-in strategy":"true"}. This record does not assert independent reproduction.

### Seed1.5-VL

Average Score: 62.9. Result ID: 69986e745698665ec4ce5c83.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78791e209131cea788a1.md>)

Recorded evaluation configuration: {"Access":"Closed-source"}. This record does not assert independent reproduction.

### Seed1.5-VL

Average Score: 62.9. Result ID: 6a06e3c947f34ce924eeca41.

[Reporting paper](<https://www.sota2.com/research/paper/6a06a8b7dbc836b87d82a1de.md>)

Recorded evaluation configuration: {"Model Category":"Proprietary Models"}. This record does not assert independent reproduction.

### GUI-Spotlight

Average Score: 62.7. Result ID: 69986e745698665ec4ce5c8c.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bdf1e209131cea8aec1.md>)

Recorded evaluation configuration: {"Access":"Open-Source"}. This record does not assert independent reproduction.

### GUI-Spotlight

Average Score: 62.7. Result ID: 6a169e16fbc373592c0f6bd9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b911e209131cea892ed.md>)

Recorded evaluation configuration: {"Initialised from":"UI-TARS-1.5-7B"}. This record does not assert independent reproduction.

### UI-Venus-1.0-72B

Average Score: 62.2. Result ID: 699722185698665ec4705c09.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ab01e209131cea849f9.md>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### Qwen2.5-VL-72B

Average Score: 62.2. Result ID: 69986e745698665ec4ce5c84.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Access":"Open-Source"}. This record does not assert independent reproduction.

### UI-Venus-Ground-72B

Average Score: 62.2. Result ID: 6a44c80873d097d845bedeee.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ab01e209131cea849f9.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"72B"}. This record does not assert independent reproduction.

### UI-TARS-1.5-7B

Average Score: 61.9. Result ID: 69986e745698665ec4ce5c8d.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

Recorded evaluation configuration: {"Access":"Open-Source","MVP Framework":false}. This record does not assert independent reproduction.

### UI-TARS-1.5-7B

Average Score: 61.9. Result ID: 6a169e16fbc373592c0f6bd5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b911e209131cea892ed.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Qwen3-VL-30B-A3B

Average Score: 61.2. Result ID: 699722185698665ec4705c00.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"reproduction":true,"design":"end-to-end"}. This record does not assert independent reproduction.

### Qwen3-VL-30B-A3B

Average Score: 61.2. Result ID: 6a44c80873d097d845bedee9.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Model Category":"General Models","Parameters":"30B-A3B"}. This record does not assert independent reproduction.

### Qwen3-30B-A3B-Instruct

Average Score: 61. Result ID: 699837d15698665ec4938e85.

[Reporting paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Parameters":"30B","Arch":"A3B","Type":"Instruct"}. This record does not assert independent reproduction.

### Qwen3-VL-32B

Average Score: 60.6. Result ID: 6a06e3c947f34ce924eeca48.

[Reporting paper](<https://www.sota2.com/research/paper/6a06a8b7dbc836b87d82a1de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Model Category":"Open-Source Models","Evaluated By Authors":"true"}. This record does not assert independent reproduction.

### Step-GUI-4B

Average Score: 60.5. Result ID: 699722185698665ec4705c0c.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Implementation](<https://github.com/stepfun-ai/gelab-zero>)

Recorded evaluation configuration: {"reproduction":true,"design":"end-to-end"}. This record does not assert independent reproduction.

### Step-GUI-4B

Average Score: 60.5. Result ID: 6a44c80873d097d845bedef8.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Implementation](<https://github.com/stepfun-ai/gelab-zero>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"4B"}. This record does not assert independent reproduction.

### GTA1-7B

Average Score: 60.1. Result ID: 699722185698665ec4705c03.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b2ba.md>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### MAI-UI-8B

Average Score: 60.1. Result ID: 699722185698665ec4705c0e.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69520ae053fdd5c36173d429.md>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### MAI-UI-8B

Average Score: 60.1. Result ID: 6a44c80873d097d845bedef9.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69520ae053fdd5c36173d429.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"8B"}. This record does not assert independent reproduction.

### OpenCUA-32B

Average Score: 59.6. Result ID: 699722185698665ec4705c02.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be61e209131cea8b2f4.md>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### UI-Venus-1.5-2B

Average Score: 59.4. Result ID: 699722185698665ec4705c10.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Implementation](<https://github.com/inclusionAI/UI-Venus>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### UI-Venus-7B

Average Score: 58.8. Result ID: 6a06e3c947f34ce924eeca44.

[Reporting paper](<https://www.sota2.com/research/paper/6a06a8b7dbc836b87d82a1de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7ab01e209131cea849f9.md>)

Recorded evaluation configuration: {"Model Category":"Open-Source Models"}. This record does not assert independent reproduction.

### Xiaomi-GUI-0-30B-A3B

Average Score: 58.7. Result ID: 6a44c80873d097d845bedefb.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

Recorded evaluation configuration: {"Model Category":"Ours","Parameters":"30B-A3B"}. This record does not assert independent reproduction.

### Trifuse+GUI-AIMA

Average Score: 58.4. Result ID: 6996005d5698665ec4366785.

[Reporting paper](<https://www.sota2.com/research/paper/698969d7c97a0b3e3b5dc31f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698969d7c97a0b3e3b5dc31f.md>)

Recorded evaluation configuration: {"Training Data Size":null,"Model Size":"3B"}. This record does not assert independent reproduction.

### GUI-AIMA

Average Score: 58.3. Result ID: 6996005d5698665ec4366784.

[Reporting paper](<https://www.sota2.com/research/paper/698969d7c97a0b3e3b5dc31f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/690988683ef7ccd74e5e5880.md>)

[Implementation](<https://github.com/sjz5202/GUI-AIMA>)

Recorded evaluation configuration: {"Training Data Size":"85K","Model Size":"3B"}. This record does not assert independent reproduction.

### Qwen3-VL-4B-Instruct

Average Score: 58.2. Result ID: 699837d15698665ec4938e83.

[Reporting paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Parameters":"4B","Type":"Instruct"}. This record does not assert independent reproduction.

### Qwen3-VL-8B-Instruct

Average Score: 58.2. Result ID: 699837d15698665ec4938e84.

[Reporting paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Parameters":"8B","Type":"Instruct"}. This record does not assert independent reproduction.

### GUI-Owl-32B

Average Score: 58. Result ID: 699722185698665ec4705c06.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b111e209131cea867fb.md>)

Recorded evaluation configuration: {"design":"end-to-end"}. This record does not assert independent reproduction.

### GUI-Owl-32B

Average Score: 58. Result ID: 699837d15698665ec4938e82.

[Reporting paper](<https://www.sota2.com/research/paper/69438a9b423ea422657294f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b111e209131cea867fb.md>)

Recorded evaluation configuration: {"Parameters":"32B"}. This record does not assert independent reproduction.

### GUI-Cursor

Average Score: 58. Result ID: 6a169e16fbc373592c0f6bd7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b911e209131cea892ed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b911e209131cea892ed.md>)

Recorded evaluation configuration: {"Initialised from":"Qwen2.5-VL-7B"}. This record does not assert independent reproduction.

### GUI-Owl-32B

Average Score: 58. Result ID: 6a44c80873d097d845bedef2.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b111e209131cea867fb.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"32B"}. This record does not assert independent reproduction.

### Trifuse+GUI-Actor

Average Score: 57.9. Result ID: 6996005d5698665ec4366783.

[Reporting paper](<https://www.sota2.com/research/paper/698969d7c97a0b3e3b5dc31f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698969d7c97a0b3e3b5dc31f.md>)

Recorded evaluation configuration: {"Training Data Size":null,"Model Size":"7B"}. This record does not assert independent reproduction.

### GUI-Owl-1.5-32B-Thinking

Average Score: 57.6. Result ID: 6a44c80873d097d845bedef7.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6997eaac09c33900fcd766ce.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"32B","Version":"1.5","Type":"Thinking"}. This record does not assert independent reproduction.

### Qwen3-VL-8B

Average Score: 57.5. Result ID: 699722185698665ec4705bff.

[Reporting paper](<https://www.sota2.com/research/paper/698c0ceec97a0b3e3b5e104f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"reproduction":true,"design":"end-to-end"}. This record does not assert independent reproduction.

### Qwen3-VL-8B

Average Score: 57.5. Result ID: 6a44c80873d097d845bedee8.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Model Category":"General Models","Parameters":"8B"}. This record does not assert independent reproduction.

### UI-TARS-72B

Average Score: 57.1. Result ID: 69986e745698665ec4ce5c85.

[Reporting paper](<https://www.sota2.com/research/paper/6938fe876cfd488761a53c83.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76a61e209131cea6e2ba.md>)

Recorded evaluation configuration: {"Access":"Open-Source"}. This record does not assert independent reproduction.

### UI-TARS-72B

Average Score: 57.1. Result ID: 6a0577c447f34ce924ec0f7b.

[Reporting paper](<https://www.sota2.com/research/paper/6a05570edbc836b87d828c27.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76a61e209131cea6e2ba.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### UI-TARS-72B

Average Score: 57.1. Result ID: 6a06e3c947f34ce924eeca49.

[Reporting paper](<https://www.sota2.com/research/paper/6a06a8b7dbc836b87d82a1de.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76a61e209131cea6e2ba.md>)

Recorded evaluation configuration: {"Model Category":"Open-Source Models"}. This record does not assert independent reproduction.

### UI-TARS-72B

Average Score: 57.1. Result ID: 6a44c80873d097d845bedeeb.

[Reporting paper](<https://www.sota2.com/research/paper/6a449f10e0224bd4f4993095.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76a61e209131cea6e2ba.md>)

Recorded evaluation configuration: {"Model Category":"GUI-specific Models","Parameters":"72B"}. This record does not assert independent reproduction.

### Qwen3-VL-8B-Thinking

Average Score: 56.7. Result ID: 699623e45698665ec464825a.

[Reporting paper](<https://www.sota2.com/research/paper/6973007bb2be0ed85a78c9c4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"Parameters":"8B","Thinking Mode":true}. This record does not assert independent reproduction.

Page 1 of 3. This is an index page, not the full collection.

- [Next index page](<https://www.sota2.com/research/sota/gui-grounding-on-osworld-g.md?metric=Average%20Score&page=2>): Follow this link to discover more pages.
