LLM Coding

Upload data, configure codebook variables, and code with one or more LLMs.

Please remember to cite our methodological paper if you use this tool:

Baranski, A., Cooper, D. J., & Lee, J. K. (2026). Are LLMs reliable coders of communicationcontent in economic experiments? NYUAD Division of Social Science Working Paper, #0115. View paper

Drop a CSV or Excel file here, or click to browse

Communication episode: the unit of analysis — a combination of messages exchanged through the same channel, or a collection of messages sent by one sender. Rows that share your chosen identifier(s) are merged into one episode.

Fully empty episodes are not sent to a model. Their original rows remain in the primary CSV with blank code cells.

No variables yet — click to define your codebook.

Click to edit codebook →

Each variable has its own aggregation method, category definition, and coded-value guidance. Per episode = one value per episode; per sender = one value per verified sender.

Download Codebook

Provide context about what the data represents and the research goals. Have a PDF with figures or tables? Use Import from PDF to convert it to text first. The placeholder is a constructed example.

Browser and package execution differ.Run Coding uses all models and runs configured below and requires their API keys. For package generation, CAT does not save an API key or the configured tuning settings; the downloaded package records only the first selected provider and model and makes one call per episode after the local script obtains a key at runtime.
1OpenAI — GPT-4.1 Mini
0.20
02
1.00
01
1024
648192
1 model × 1 run = 1 calls/episode

CAT Privacy Notice

How CAT processes research data, API keys, optional analytics, and contact messages.

Last updated: August 20, 2026

Who operates CAT

CAT is operated by the Social Science Experimental Laboratory at NYU Abu Dhabi. Privacy, access, correction, and deletion requests may be sent to jkl499@nyu.edu.

Optional usage analytics

CAT asks before collecting optional analytics. If you accept, CAT records a persistent browser session identifier, public IP address, best-effort country, region and city, browser user agent, referring page, event time, selected provider and model names, and counts describing the configured task. CAT obtains the public IP through ipify and sends it to ip-api.com to obtain an approximate location. These services receive the IP address needed to answer those requests.

If you reject optional analytics, CAT sends only an anonymous visit event. The application does not look up or store your location, browser identifier, user agent, referrer, or task configuration for that event. As with any website, university infrastructure and reverse- proxy access logs may process connection information for security and operations.

Your choice is stored in this browser as cat_analytics_consent. Use the button below to make a new choice for future events.

Demonstration video

The demonstration video is delivered on demand from CAT's public GitHub repository. Your browser does not request the video until you press play. GitHub receives the connection information needed to deliver the video, which may include your IP address and browser request metadata.

Research data and API keys

CAT does not store API keys in its database, browser storage, generated packages, or application logs. Keys are held in memory for the requested operation. During browser coding, the selected LLM provider receives the experimental instructions, coding manual, communication episodes, and selected context needed for the task.

Uploaded datasets and generated results use temporary server-side working files. Project reset and successful dataset replacement request cleanup, and scheduled cleanup removes working directories after the 24-hour threshold. The browser also keeps the current project and dataset locally so work can be restored; resetting the project clears that browser copy.

Contact messages and retention

Contact-form submissions contain the name, email address, title, message, status, and time supplied for responding to the request. Operational analytics and contact messages are currently retained until administratively deleted. You may request access or deletion at the email address above. CAT does not sell personal information.

Your responsibilities

Researchers are responsible for confirming that data sent through CAT complies with participant consent, institutional review requirements, data-use agreements, and applicable law. Sensitive data should be de-identified or used only with an institutionally approved LLM provider.