Quotas & Limits

This document contains the current restrictions and usage quotas on use of AutoML Natural Language. This page will be updated to reflect any changes to these restrictions and usage quotas.

These quotas apply to each AutoML Natural Language project, and are shared across all applications and IP addresses using a given project.

You can edit your quotas up to their maximum values by selecting EDIT QUOTAS from the Quotas page of the Google Cloud Platform Dashboard. Edit the quotas for the Cloud AutoML API. To request an increase above the maximum quota, edit your quota with your requested increase and justification then submit your update. You will be notified when your request has been received. You might be contacted for more information regarding your request. After your request has been reviewed, you will be notified whether it has been approved or denied.

AutoML Natural Language has quotas associated with creating models and with making prediction requests to the models.

Modeling quotas

Activity Default quota
Model creation requests per minute 10
Concurrent model training requests 5
Deployed models 30

Request quotas

These quotas apply collectively to all deployed models.

Activity Default quota
Prediction requests per minute per project (online) 600
Non-prediction requests per minute per project 600
For entity extraction, the number of active batch requests per project 10
For classification and sentiment analysis, the number of active batch requests per project 5

Limits

Character counts assume UTF-8 characters.

Type of limit Classification Entity extraction Sentiment analysis
Training items 10 to 1,000,000 50 to 100,000 10 to 100,000
Labels per dataset 2 to 5,000 1 to 100 2 to 11
Length of label name 1 to 32 1 to 32 Integer from 0 to 10
Length of annotated span N/A 1 to 100 characters N/A
Training items per label 10 to 1,000,000 100 to 100,000 10 to 100,000
Training item size 10MB
5,000,000 characters
128 KB (text); 20MB (PDF)
10 to 300,000 characters (text)
128 KB (text); 2MB (PDF)
60,000 characters
Item sent for prediction 128 KB (text); 2MB (PDF)
60,000 characters
20MB 128 KB (text); 2MB (PDF)
60,000 characters
Items per batch request 10,000 10,000 10,000