| --- |
| language: |
| - en |
| - de |
| - fr |
| - es |
| - it |
| - nl |
| license: cc-by-4.0 |
| size_categories: |
| - 1K<n<10K |
| task_categories: |
| - image-to-text |
| - visual-question-answering |
| - object-detection |
| pretty_name: Pointerbench |
| tags: |
| - gui-grounding |
| - computer-use |
| - benchmark |
| - screenshots |
| - synthetic-data |
| - spreadsheets |
| - text-grounding |
| - professional-apps |
| --- |
| |
| # Pointerbench |
|
|
| Paper: [PointerBench: A Diagnostic Benchmark for Fine-Grained GUI Grounding](https://huggingface.co/papers/2610.31274) |
|
|
| Pointerbench is a small GUI grounding benchmark suite for computer-use models. |
| Each example has one screenshot, one instruction, target geometry in absolute |
| pixels, and a binary evaluation rule. |
|
|
| Links: |
|
|
| - GitHub: https://github.com/warmwindOS/pointerbench |
| - Blog post: https://about.warmwind.com/pointer-bench/ |
| - Add your model to the official benchmark leaderboard: https://warmwind.com/contact |
|
|
| The suite has three subsets: |
|
|
| | Subset | Examples | What it tests | |
| | --- | ---: | --- | |
| | `pointerbench-sheets` | 500 | Spreadsheet cells, colors, headers, edges, corners, and relative positions | |
| | `pointerbench-text` | 500 | Words, characters, punctuation, caret positions, chrome text, text bounding boxes, and invoice fields | |
| | `pointerbench-pro` | 500 | Icons, text, and mixed GUI targets across 100 professional applications | |
|
|
| ## Leaderboard |
|
|
| Scores are percentages. The table is ranked by the reported public average across Sheets, Text, and Pro. |
|
|
| | Rank | Model | Sheets | Text | Pro | Average | |
| | ---: | --- | ---: | ---: | ---: | ---: | |
| | 1 | `anthropic/claude-fable-5` | 96.0 | 58.4 | 80.8 | 78.7 | |
| | 2 | `warmwind/pointer-1.6 (agentic)` | 87.8 | 66.8 | 76.8 | 77.1 | |
| | 3 | `anthropic/claude-sonnet-4.6` | 82.6 | 47.8 | 77.4 | 69.3 | |
| | 4 | `anthropic/claude-opus-4.8` | 82.2 | 44.4 | 80.2 | 68.9 | |
| | 5 | `warmwind/pointer-1.6` | 77.0 | 49.2 | 76.8 | 67.7 | |
| | 6 | `openai/gpt-5.6-sol-pro` | 92.0 | 32.0 | 72.0 | 65.3 | |
| | 7 | `openai/gpt-5.6-sol` | 92.0 | 30.0 | 74.0 | 65.3 | |
| | 8 | `anthropic/claude-opus-4.7` | 73.8 | 36.2 | 75.4 | 61.8 | |
| | 9 | `openai/gpt-5.5` | 72.4 | 31.4 | 75.8 | 59.9 | |
| | 10 | `openai/gpt-5.6-terra-pro` | 52.0 | 20.0 | 74.0 | 48.7 | |
| | 11 | `openai/gpt-5.6-terra` | 50.0 | 22.0 | 70.0 | 47.3 | |
| | 12 | `openai/gpt-5.4` | 63.2 | 25.0 | 50.4 | 46.2 | |
| | 13 | `openai/gpt-5.6-luna` | 56.0 | 20.0 | 60.0 | 45.3 | |
| | 14 | `openai/gpt-5.6-luna-pro` | 54.0 | 18.0 | 62.0 | 44.7 | |
| | 15 | `openai/gpt-5` | 35.2 | 13.6 | 20.2 | 23.0 | |
| | 16 | `x-ai/grok-4.20-multi-agent` | 10.6 | 1.4 | 10.0 | 7.3 | |
| | 17 | `moonshotai/kimi-k2.5` | 10.8 | 3.4 | 3.0 | 5.7 | |
| | 18 | `openai/gpt-5-mini` | 6.0 | 1.4 | 5.8 | 4.4 | |
| | 19 | `moonshotai/kimi-k2.6` | 5.6 | 2.8 | 2.8 | 3.7 | |
| | 20 | `qwen/qwen3.7-plus` | 5.8 | 0.6 | 3.0 | 3.1 | |
| | 21 | `qwen/qwen3.6-flash` | 4.2 | 1.2 | 2.6 | 2.7 | |
| | 22 | `minimax/minimax-m3` | 3.6 | 0.4 | 3.6 | 2.5 | |
| | 23 | `qwen/qwen3.6-plus` | 3.0 | 0.6 | 3.6 | 2.4 | |
| | 24 | `qwen/qwen3-vl-235b-a22b-thinking` | 2.6 | 0.4 | 3.2 | 2.1 | |
| | 25 | `qwen/qwen3.5-9b` | 2.0 | 0.2 | 2.6 | 1.6 | |
| | 26 | `google/gemini-3.5-flash` | 3.4 | 0.6 | 0.6 | 1.5 | |
| | 27 | `openai/gpt-5-nano` | 2.0 | 0.2 | 1.6 | 1.3 | |
| | 28 | `moonshotai/kimi-k2.7-code` | 0.4 | 1.8 | 0.8 | 1.0 | |
| | 29 | `x-ai/grok-4.3` | 1.2 | 0.0 | 0.0 | 0.4 | |
| | 30 | `google/gemini-3.1-pro-preview` | 0.0 | 0.2 | 0.4 | 0.2 | |
| | 31 | `google/gemini-3.1-pro-preview-customtools` | 0.0 | 0.4 | 0.0 | 0.1 | |
| | 32 | `stepfun/step-3.7-flash` | 0.0 | 0.0 | 0.2 | 0.1 | |
| | 33 | `x-ai/grok-build-0.1` | 0.0 | 0.0 | 0.0 | 0.0 | |
| | 34 | `xiaomi/mimo-v2.5` | 0.0 | 0.0 | 0.0 | 0.0 | |
| | - | Center-click baseline | 0.4 | 0.2 | 0.2 | 0.3 | |
|
|
| To add your model to the official benchmark leaderboard, contact https://warmwind.com/contact. |
|
|
| All images are synthetic 1024x768 PNG screenshots. The datasets contain no |
| scraped user data and no PII. |
|
|
| ## Layout |
|
|
| Each subset is self-contained: |
|
|
| ```text |
| pointerbench-sheets/ |
| data/test/metadata.jsonl |
| data/test/0000.png |
| eval.py |
| README.md |
| REPRODUCE.md |
| |
| pointerbench-text/ |
| data/test/metadata.jsonl |
| data/test/0000.png |
| eval.py |
| README.md |
| REPRODUCE.md |
| |
| pointerbench-pro/ |
| data/test/metadata.jsonl |
| data/test/0000.png |
| eval.py |
| README.md |
| REPRODUCE.md |
| ``` |
|
|
| ## Schema |
|
|
| Each metadata row includes: |
|
|
| ```json |
| { |
| "file_name": "0000.png", |
| "id": "pbs_0000", |
| "instruction": "Click cell E11.", |
| "bbox": [x1, y1, x2, y2], |
| "point": [x, y], |
| "answer_type": "point", |
| "eval": {"type": "point_in_bbox", "bbox": [x1, y1, x2, y2]}, |
| "data_type": "cell", |
| "category": "cell_ref", |
| "image_size": [1024, 768] |
| } |
| ``` |
|
|
| Point tasks are correct when the predicted point lands inside the target bbox. |
| Bbox tasks, used in Pointerbench-Text, use an asymmetric overlap rule: a hit |
| requires the ground truth to be almost fully covered (coverage >= 0.90) and the |
| prediction to stay reasonably tight around it (precision >= 0.70). This |
| penalises predictions that cut off part of the target far more than predictions |
| that wrap it with some margin. |
|
|
| ## Evaluation |
|
|
| Run the scorer inside a subset folder: |
|
|
| ```bash |
| python eval.py --predictions preds.jsonl |
| ``` |
|
|
| Predictions are JSONL rows with an `id` and either a `point` or `bbox`, depending |
| on `answer_type`. |
|
|
| Recommended inference prompt: |
|
|
| ```bash |
| python eval.py --show-system-prompt |
| ``` |
|
|
| ```text |
| You are evaluating Pointerbench, a GUI grounding benchmark. You will receive one 1024x768 screenshot and one task instruction. Use absolute pixel coordinates with origin at the top-left of the image. Do not return normalized coordinates. Do not crop or resize the coordinate frame. For point tasks, return JSON like {"point": [x, y]}. For bounding-box tasks, return JSON like {"bbox": [x0, y0, x1, y1]}. |
| ``` |
|
|
| You can edit the prompt for your inference stack. Keep the 1024x768 absolute |
| pixel coordinate frame fixed, and report any image resizing or multi-step zoom |
| strategy with your results. |
|
|
| See each subset README for the exact distribution, schema details, and examples. |
|
|
| ## License |
|
|
| Dataset images and annotations are released under CC BY 4.0. The included |
| evaluation scripts are released under MIT. |
