Saturday, May 28, 2011 6:25 AMOwner
We released an update to the Microsoft Translator service today, changing the previously enforced throughput limits. For the vast majority of our API users, this change has increased the throughput limits.
- The limit of the number of requests per minute is removed.
- An hourly and a daily quota limit the number of characters Microsoft Translator will accept.
- The request size remains limited at 10000 characters.
The quota will adjust to ensure fairness among all users of the service.
You will receive an error message "AppId is over the quota" if your appid exceeded the hourly or daily quota. You will receive an error "IP is over the quota" if your IP address exceeded the quota.If you have signed a commercial use license with us, please contact us at email@example.com to ensure the optimal service for your app, and let us know the appid you are using.As a side note: the Microsoft Translator API is here to stay.Chris WendtMicrosoft Translator
- Changed Type Microsoft TranslatorOwner Tuesday, June 07, 2011 11:39 PM
Sunday, May 29, 2011 9:12 AMWhat are the hourly and monthly limits?
Sunday, May 29, 2011 10:19 PMOwner
Commercial use of the Microsoft Translator API requires a commercial license. The throughput limits will be whatever we agree on, considering your application's needs.
The throughput limits for non-commercial use will vary based on service utilization, to ensure fairness.
Friday, July 15, 2011 9:49 PM
Does using the "translate array" method, instead of just the "translate" method, affect your quota in any way?
For example, does submitting one translate_array request containing 10 sentences, count the same as 10 translate requests, on the same sentences?
Monday, July 18, 2011 3:00 PMOwner
All methods are counted the same way: the sum of the request sizes is the significant value. You can economize on the network latency by sending distinct text segments in a single call using the ...Array() methods.
Note that all elements in TranslatreArray() are translated from the same language and to the same language, even when using automatic language detection.