RPC-Bench Leaderboard
How to Submit
- Fork this repository.
- Create a new branch for your submission.
- Add your submission folder under
submissions/<organization>__<model>__<input_config>/. - Open a Pull Request with the new submission folder.
Submission Directory Requirements
Each submission directory must contain the metadata and predictions for one model/input configuration pair:
<organization>__<model>__<input_config>/
metadata.yaml
predictions.jsonl
generation_config.json # optional, recommended
artifacts/ # optional logs or prompt notes
Use URL-safe directory names. Replace spaces, slashes, and special characters
with hyphens; keep input_config as TEXT or VISUAL.
metadata.yaml
model_name: "My Model"
organization: "My Org"
model_url: https://... # optional work link: paper, GitHub, model card, etc.
date: "2026-06-17" # model release date, not submission date
split: test
input_config: TEXT # TEXT or VISUAL
predictions.jsonl
Each line must be one JSON object:
{
"id": "paper-id",
"part_idx": 1,
"question": "question text",
"category": "category",
"gen_answer": "model answer"
}
part_idx is the question index in the current paper's qa_pairs list (1 for the first item). category must match the corresponding item in test.json.
Validation Rules
Your submission will be validated before evaluation. To pass:
metadata.yamlmust includemodel_name,organization,date,split, andinput_config.model_urlis optional.dateis the model release date, not the submission date.splitmust betest.input_configmust beTEXTorVISUAL.predictions.jsonlmust contain exactly one line for every QA item intest.json.part_idxis the question index in the current paper'sqa_pairslist (1for the first item).id,part_idx,question, andcategorymust exactly match the benchmark item.gen_answermust be a string.- For
Claim_Verification,gen_answermust be exactlyTrueorFalse.
Submission Process
- Open PR: add your folder under
submissions/<organization>__<model>__<input_config>/. - Fix issues: if validation fails, update the PR with corrected files.
- Review: once validation passes, a maintainer reviews the submission.
- Evaluate: maintainers run the official evaluator in a controlled local environment.
- Import: accepted aggregate results are imported to the leaderboard.
Leaderboard
28 entries
| Rank | Model | Organization | Input Config | Conciseness | Correctness | Completeness | F1-like | Informativeness | Date |
|---|---|---|---|---|---|---|---|---|---|
| 1 | GPT-5 | OpenAI | TEXT | 54.93 | 69.10 | 67.33 | 68.20 | 37.46 | 2025-8-7 |
| 2 | GPT-5.2 | OpenAI | TEXT | 53.81 | 66.84 | 64.03 | 65.40 | 35.19 | 2025-12-11 |
| 3 | GPT-5 | OpenAI | VISUAL | 61.47 | 58.90 | 55.34 | 57.07 | 35.08 | 2025-8-7 |
| 4 | Gemini-2.5-Pro | TEXT | 54.87 | 62.65 | 59.03 | 60.79 | 33.35 | 2025-3-25 | |
| 5 | Gemini-3-Pro | TEXT | 52.81 | 62.69 | 60.28 | 61.46 | 32.46 | 2025-11-18 | |
| 6 | DeepSeek-V3.2 | DeepSeek-AI | TEXT | 56.31 | 58.73 | 55.19 | 56.91 | 32.04 | 2025-12-1 |
| 7 | GPT-5.2 | OpenAI | VISUAL | 56.43 | 56.75 | 52.82 | 54.72 | 30.88 | 2025-12-11 |
| 8 | DeepSeek-V3.1 | DeepSeek-AI | TEXT | 54.76 | 57.85 | 54.85 | 56.31 | 30.84 | 2025-8-21 |
| 9 | GLM-4.6V | Z.ai | VISUAL | 64.55 | 47.32 | 43.43 | 45.29 | 29.23 | 2025-12-8 |
| 10 | GLM-4.7 | Z.ai | TEXT | 54.34 | 54.36 | 51.75 | 53.02 | 28.81 | 2025-12-22 |
| 11 | GLM-4.5V | Z.ai | VISUAL | 59.44 | 48.79 | 43.62 | 46.06 | 27.38 | 2025-8-11 |
| 12 | gemini-3-pro | VISUAL | 50.22 | 56.06 | 52.69 | 54.32 | 27.28 | 2025-11-18 | |
| 13 | GLM-4.5 | Z.ai | TEXT | 43.41 | 58.95 | 59.54 | 59.24 | 25.72 | 2025-7-28 |
| 14 | gemini-2.5-pro | VISUAL | 51.71 | 48.39 | 45.59 | 46.95 | 24.28 | 2025-3-25 | |
| 15 | Claude-Sonnet-4 | Anthropic | TEXT | 41.37 | 58.53 | 58.44 | 58.48 | 24.19 | 2025-5-23 |
| 16 | Qwen3 | Alibaba | TEXT | 41.44 | 55.88 | 56.64 | 56.26 | 23.31 | 2025-7-21 |
| 17 | Claude-Sonnet-4.5 | Anthropic | TEXT | 31.02 | 64.31 | 64.97 | 64.64 | 20.05 | 2025-9-30 |
| 18 | Claude-Sonnet-4.5 | Anthropic | VISUAL | 31.95 | 55.35 | 54.45 | 54.89 | 17.54 | 2025-9-30 |
| 19 | Claude-Sonnet-4 | Anthropic | VISUAL | 31.63 | 54.16 | 53.32 | 53.74 | 16.99 | 2025-5-23 |
| 20 | HippoRAG2 | The Ohio State University | TEXT | 45.77 | 33.13 | 27.88 | 30.28 | 13.86 | 2025-6-19 |
| 21 | MemoRAG | Peking University & Hong Kong Polytechnic University | TEXT | 51.31 | 24.19 | 19.10 | 21.35 | 10.96 | 2025-4-9 |
| 22 | VdocRAG | NTT Corporation & Tohoku University | VISUAL | 61.54 | 21.17 | 13.88 | 16.77 | 10.32 | 2025-4-14 |
| 23 | VisRAG | Tsinghua University & ModelBest Inc. | VISUAL | 39.90 | 26.24 | 23.63 | 24.87 | 9.92 | 2025-3-2 |
| 24 | Raptor | Stanford University | TEXT | 36.47 | 25.28 | 20.82 | 22.84 | 8.33 | 2024-1-31 |
| 25 | Monkey | Huazhong University of Science and Technology | VISUAL | 54.61 | 17.08 | 11.27 | 13.58 | 7.41 | 2024-8-26 |
| 26 | Docopilot | Shanghai AI Laboratory | VISUAL | 39.31 | 18.31 | 17.12 | 17.69 | 6.96 | 2025-7-19 |
| 27 | Qwen3 | Alibaba | VISUAL | 22.64 | 20.17 | 20.14 | 20.16 | 4.56 | 2025-7-21 |
| 28 | DocOwl2 | Alibaba | VISUAL | 50.19 | 11.75 | 6.66 | 8.50 | 4.27 | 2024-9-9 |