Upload a dataset (CSV, JSON, XLSX, TXT), add your NVIDIA NIM or IU REALLMS API key,
write a prompt, and let a language model annotate every row or chunk.
Your key never leaves your browser โ all calls go directly to the API.
โ Upload File
๐
Drop a file here, or click to browse
CSV ยท JSON ยท XLSX ยท TXT ยท and more
Select which column contains the text the LLM should read.
โก API Configuration
Your key is used only for direct calls from your browser to the provider.
It is never sent to or stored on this website.
Use {text} as a placeholder โ it will be replaced with each row's content.
For CSV/XLSX, you can also reference specific columns with {column_name}.
โฃ Options
Download:
Processingโฆ0 / 0
Startingโฆ
๐ฅDownload Annotated Dataset
Your file with model annotations is ready for export.
Live Annotation Feed
How this works
Your file is parsed locally in the browser. Each row (or chunk of text) is sent
one-by-one to the LLM API using your key. The model's response is saved back
into a new column you can export as CSV or JSON.
Provider endpoints
Both providers use an OpenAI-compatible /v1/chat/completions interface.
NVIDIA NIMnvai-api
integrate.api.nvidia.com/v1
IU REALLMSIU
reallms.rescloud.iu.edu/direct/v1
Session stats
Rows loadedโ
Annotatedโ
Errorsโ
File typeโ
๐ Privacy: Your API key and file data never leave your browser
except as direct calls to your chosen provider.
โ ๏ธ Rate limits: Use the delay setting to stay within free-tier
limits (NVIDIA free tier: ~5 req/s).