Ctrlk
Visit our websitePricingBlogPlaygroundAPI Docs
For the complete documentation index, see llms.txt. This page is also available as Markdown.

LLM Assisted Labeling

Supported labeling types: Span labeling, row labeling.

LLM Assisted Labeling allows you to use models from OpenAI, Azure OpenAI, Anthropic, Gemini, and Cohere to assist with labeling. This helps make labeling easier across different use cases.

Providers and model support

Provider name
Model

OpenAI

gpt-5.4, gpt-5.4-mini, gpt-5.1, gpt-4.1, gpt-4.1-mini, gpt-4o, gpt-4o-mini

Azure Open AI

Anthropic

claude-sonnet-5, claude-opus-5, claude-sonnet-4-6, claude-haiku-4-5, claude-sonnet-4-5

Gemini

gemini-3.5-flash, gemini-2.5-pro, gemini-2.5-flash, gemini-flash-latest, gemini-pro-latest

Cohere

command-a-03-2025, command-r-plus-08-2024, command-r-08-2024, command-r7b-12-2024

Custom

Support for any LLM Provider without requiring additional coding for Row Labeling.

Quick guide

After creating the project, enable the ML-assisted labeling extension and select LLM Assisted Labeling as the provider. The following fields will be available:

Image of Empty Row Based Project

For custom providers, refer to the dedicated page.

  1. LLM provider: Select a supported LLM provider.

  2. Target text: Select Input column(s) that will be used as context.

  3. Target question: Select a question to be answered.

  4. System prompt: Define the model’s behavior and context.

  5. User prompt: Defines the task for the model.

  6. API key: Add the LLM provider secret key.

  7. For Azure OpenAI

    1. API version: Enter the API version from your Azure OpenAI.

    2. API base URL: Enter the base URL for your Azure OpenAI API model.

    3. Model deployment: Enter the deployment model name from Azure OpenAI.

  8. Advanced settings

    1. Top P: Limits predictions to the smallest set with a cumulative probability of P.

    2. Temperature: Controls randomness; lower values make responses more predictable.

    3. Maximum tokens: Limits the length of the generated response.

    4. Model name: The specific version of the language model.

For guidance, you can refer to our prompt examples: row labeling and span labeling.

After completing the configuration, click Predict labels to generate labeling suggestions.

Image of Prediction Result

A 429 error indicates rate limits from the LLM provider. Check your API usage and limits.

Last updated