Run your evals. Skip the setup.
Spend less time on evals. More on NLP research.
Upload a spreadsheet, compare models, and get results ready for your paper.
Free to use. Bring your own key, or use free models.
Spreadsheet to paper in six steps
- 1
Upload an Excel or CSV file.
- 2
Map columns to inputs and the answer key.
- 3
Write a prompt using your columns, like {sentence}.
- 4
Pick models, free or paid.
- 5
Run with retries and live progress.
- 6
Export to Excel, LaTeX, and a reproducibility config.
Compare models in one run
Run several models on the same data and see scores side by side.
Built for free tiers
Bring your own key. Requests are paced to fit free limits.
Metrics that fit research
Accuracy, F1, and more, with tokenisation that works for Burmese, Thai, Chinese, Japanese, and Korean.
Paper-ready output
A LaTeX table and an Excel workbook with every score.
Reproducible
Every run records models, prompt, seed, and a data hash, so reviewers can repeat it.
Your data stays yours
Keys are encrypted, and you can delete everything anytime.
Open source at the core
The engine is open source (Apache 2.0) and citable. Prefer the command line? Use it directly.
pip install paperlab-space
paperlab-space init my_data.xlsx
paperlab-space run run.yaml