Skip to content

Reference

Transformation LLM

Transforms data using an LLM prompt. Combines the power of a language model with structured data transformation.

2 min

Transforms data using an LLM prompt. Combines the power of a language model with structured data transformation.

Parameters

ParameterTypeRequiredVariableDescription
inputArraydynamic valueYesYesArray of data to transform using the AI model.
itemVariabletextNoNoVariable name representing each item in the prompt template. (Default: "item")
promptTemplatetextYesYesTransformation instructions applied to each item. Supports {{item}}.
systemPrompttextNoYesSystem instructions sent before the transformation prompt.
model.providertextYesNoAI model provider (e.g. openai, anthropic).
model.modeltextNoNoModel identifier to use (e.g. gpt-4o-mini).
parameters.temperaturenumberNoNoModel creativity (0 = deterministic, 1 = creative). (Default: 0.7, min 0, max 2)
parameters.maxTokensnumberNoNoMaximum number of tokens per individual response. (Default: 500, min 1, max 128000)
batchSizenumberNoNoNumber of items processed per batch. (Default: 10, min 1, max 100)
parallelismnumberNoNoNumber of batches processed in parallel. (Default: 3, min 1, max 10)
outputFormatchoice (text, json, structured)NoNoOutput format per item: plain text, JSON, or typed structure. (Default: "text")
outputFieldtextNoNoField name containing the LLM result in each item. (Default: "llmResult")
retryOnErrorbooleanNoNoAutomatically retry on model errors. (Default: true)
maxRetriesnumberNoNoMaximum number of retries on error. (Default: 2, min 0, max 5)
outputVariabletextNoNoOutput variable name containing the transformed items.

Parameters marked Variable = Yes accept the {{blockName.field}} syntax.

Output

Output variable : llmTransformResult

{
  "success": false,
  "results": [],
  "totalItems": 0,
  "tokensInput": 0,
  "tokensOutput": 0,
  "totalTokens": 0,
  "durationMs": 0
}

Example

Normalize addresses with an LLM.

Input :

{"address": "12 rue victor hugo 75001 paris"}

Output :

{"content": "12 Rue Victor Hugo, 75001 Paris, France"}

Tip Useful for semantic transformations impossible with code (spell correction, normalization, classification).

Did this page help you achieve your goal?