---
sourceDocument: Australia Enable AI
sourceDocumentLink: https://servicenow-prod.fluidtopics.net/r/pt-BR/intelligent-experiences

 Release :

    - australia

ft:locale :

    - pt-BR

ft:publication_title :

    - Australia Enable AI

ft:clusterId :

    - platai

bundleId :

    - platai

workflow :

    - Platform


---

# Create an NLU model from a CSV file

# Create an NLU model from a CSV file {#ariaid-title1}

* Versão de lançamento: Australia
* 
* Atualizado 12 de mar. de 2026
* 
* ![](https://www.servicenow.com/docs/portal-asset/ico-clock) 1 min. de leitura

Upload a CSV or XLSX (Excel Workbook) file containing utterances and their intents to
create a Natural Language Understanding (NLU) model. Use this method to quickly create models
from your data or other exported models.

## Antes de Iniciar

* Make sure that the NLU Workbench plugin, NLU Workbench - Core plugin, and Predictive Intelligence plugin are all installed and activated on your instance.
* You can create NLU models for Virtual Agent and AI Search.
* Role required: admin or nlu_admin
{#create-nlu-model-csv__ul_shl_vvs_tgb}

## Por Que e Quando Desempenhar Esta Tarefa

In this example procedure, you're building an NLU model to help Virtual Agent understand user requests regarding calendars.  
In a CSV file, you've listed the intents and utterances in two columns.Figura 1. Sample CSV setup

Note the following for creating NLU models by CSV import:

* A model needs at least 1 intent with a minimum of 5 training utterances in each intent. For optimum performance, aim to have 15 training utterances per intent.
* Utterances should not contain a comma.
* Importing with a CSV file does not preserve entities. Make sure to annotate utterances as needed after importing.

## Procedimento

1. Set your scope to the application scope you want for your new model.
2. Navigate to AllNLU WorkbenchModels.  
   The Virtual Agent tab opens by default.
3. Select the tab for the type of model you want to create, such as AI Search.
4. Select the Create new model button.
5. In the How do you want to create your model? window, select Import data from a CSV.
6. In the Add some details window, add the Name and Short description for the model.  
   In this example scenario, you enter <kbd class="ph userinput">Calendar Model</kbd> for the name and <kbd class="ph userinput">Model for answering and performing calendar requests</kbd> for the short description.
7. Select the language and purpose from the drop-down lists.  
   In this example scenario, you select <kbd class="ph userinput">English</kbd> and <kbd class="ph userinput">Virtual Agent</kbd>.
8. Click Next.
9. On the Import CSV screen, click Select file.  
10. Choose the CSV or XLSX (Excel Workbook) file from the pop-up.
11. Select Next.  
    Your model starts building. After completion, select View model to open the model details page.

## O que Fazer Depois

Add intents and training utterances to continue building the model. Add entities and vocabulary to help the model understand inputs from your users. For more information, see [Build and train your model](https://servicenow-prod.fluidtopics.net/EstdFaD6Z~spTHTfjmvTdg "After creating a model, build the model's content by adding intents, entities, vocabulary, and test set utterances. Your NLU model content determines how the model responds to user inputs.").

Add test utterances and intents to build the model's default test set. For more information, see [Test set creation and management](https://servicenow-prod.fluidtopics.net/WacY_xcSz2_eyKswVGmsBA "Use the default test set of your NLU model to test the model's performance and accuracy. Manage your test set over time by building or updating its content in the NLU Workbench.").