Quotas & Limits

This document contains current API restrictions and usage quotas on use of the Translation API. This page is updated when there are changes to these restrictions and usage quotas.

These quotas apply to each Translation API developer project and each user, and are shared across all applications and API versions used by a given developer project.

You can edit your quotas up to their maximum values by using the Google Cloud Platform Dashboard. To request an increase above the maximum for a quota, submit a request from the Google Cloud Platform Console page.

Content quotas

Content is provided as text strings to the Translation API. Both pricing and usage are based on character count. All characters sent to the Translation API count toward these quotas, including whitespace characters.

Daily quotas reset at midnight Pacific Time.

The Translation API is optimized for translation of short requests. The recommended maximum length for each request is 2K code points. The Translation API rejects very large requests and yields a 400 INVALID_ARGUMENT error) regardless of the available quota.

Character-based quota applies to the translateText and detectLanguage methods. The hard limit for requests is 30K Unicode characters (code points).

Content Quota	Default	Maximum	Duration	Applies to
General model characters per day	1 billion	unlimited	day	project
General model characters per 100 seconds per project	1,000,000 characters	10,000,000 characters	100 seconds	project
General model characters per 100 seconds per project per user	100,000 characters	10,000,000 characters	100 seconds	project and user

Note: Quota daily limits are updated once a day, which means new limits may take up to 24 hours to become effective.

Note: User limit - Number of requests an end user using an API consumer's client application can make in a time window (100 seconds). When you use a service account, the service account's identity is the "user" for quota purposes. For requests using an API key, the client IP address is used to enforce this limit.

If you exceed your quota, Translation v2 returns a 403 error. The error message says Daily Limit Exceeded if you exceeded the daily limit or User Rate Limit Exceeded if you exceeded either of the General model characters per 100 seconds quotas.

To view or change usage limits for your project, or to request an increase to your quota, do the following:

1. If you don't already have a billing account for your project, then create one.
2. Visit the Enabled APIs page of the API library in the GCP Console, and select an API from the list.
3. To view and change quota-related settings, select Quotas. To view usage statistics, select Usage.

Request quotas

In addition to the content quotas shown above, Cloud Translation enforces quotas on the usage of the API.

Discover supported languages requests

For Cloud Translation v2 and v3beta1, there is a limit on how often you can request the list of supported languages for translation.

This quota does not apply to text sent for translation. It applies only to requests for the list of supported languages.

Quota	Default	Maximum	Duration	Applies to
Requests for the list of supported languages	1000 requests	1000 requests	100 seconds	project

v3beta1

For Cloud Translation v3beta1, there is a limit to the number of requests you can make. This quota applies to all Cloud Translation v3beta1 method calls except translateText, detectLanguage, getSupportedLanguages. translateText falls under General model characters per 100 seconds for general models (see above), and AutoML model characters per minute for AutoML custom models. detectLanguage and getSupportedLanguages are covered by General model characters per 100 seconds and Requests for the list of supported languages per 100 seconds quotas, respectively (see above). Whereas, batchTranslationText falls under v3 batch translation characters using general models per project per day for general models, and v3 batch translation characters using AutoML custom models per project per day for AutoML custom models. These two quotas are set to unlimited by default and reset at Pacific Time midnight.

This quota does not apply to requests sent to Cloud Translation v2.

Activity	Default quota	Duration	Applies to
v3 default requests per project	6000	60 seconds	project
v3 default requests per project per user	unlimited	60 seconds	project and user
AutoML model characters per project	100,000	60 seconds	projects
AutoML model characters per project per user	unlimited	60 seconds	project and user
Batch translation characters using general models per project	unlimited	1 day	project
Batch translation characters using AutoML custom models per project	unlimited	1 day	project

Models (v3beta1 only)

When a custom or general model is used, content quota is checked against the project ID associated with the model. The request quotas are checked against the project associated with the request.

For example consider this translate text request:

POST https://translation.googleapis.com/v3beta1/projects/project-id-1/locations/us-central1:translateText
{
"Model":"projects/project-id-2/locations/us-central1/models/model-id",
    "sourceLanguageCode": "en",
    "targetLanguageCode": "ru",
    "contents": ["Dr. Watson, please discard your trash."]
}

The request quotas are checked against project-id-1 but content quota is checked against project-id-2.