Training my model costs 300 USD it says. 300 USD to train my model and evaluate your product? Without seeing any results beforehand? Are you out of your mind? First show me what you can do, THEN I will consider paying for your services. You've got guts, I have to admit!

The current Maximum Request Size (characters) is 10,000 characters. It is recommended to split the content to smaller chunks, which can be done in a plain text, but this brings complexity for HTML input with nested structure. For our purpose increasing somewhere around 30K should help.

Our project requires Latin American Spanish. Given that the majority of Spanish speakers are from Latin America it would be great to have a translation option for this region.

Please find attached file of the mistakes I could found it, for example, you made opposite of Kurdish(central) and Kurdish(northern) alphabet in outlook also in Kurdish swift keyboard there are some wrong letters that there's no such letter in kurdish central alphabet

Please find attached file of the mistakes I could found it, for example, you made opposite of Kurdish(central) and Kurdish(northern) alphabet in outlook also in Kurdish swift keyboard there are some wrong letters that there's no such letter in kurdish central alphabet

Match the quotas of Google Translation API.

This will expand Azure's addressable market and raise awareness of Azure to more of Microsoft's target users: developers and businesses.

The Google Translation API allows 30K characters per request and allows 30K array elements. Because of this, many non-traditional use cases, such as localizing mobile apps (which often contain hundreds/thousands of discrete strings in need of translation) are far better with Google Translate because of Azure's restrictive quotas.

We have had several customers request Azure but ultimately choose Google for the translation engine because of these limits.

My project is using Translation Services which supports Korean, but I notice there is no Korean support for the Transliteration feature. Would love to see it added in as part of the Translation API!

I think this is a bug. The v3 API shouldn't translate/convert punctuation into Chinese characters when translating en to zh. For example (http://abc.com) becomes （http：//abc.com）, which ruins the markdown syntax.
It should leave the punctuation as is or at least add a parameter in the API for users to decide.

I know there is a class=notranslation but that's not flexible for plain text and adds complexity as well.

The languages API shows me the Azure does not support Traditional Chinese in Dictionary class, could you support this part? or when ?

It will be great to have an option for fuzzy search to find the misspelled words by users. If API was called by the word "boks" for lookup from English, we will receive fuzzy results, which will include "books", instead of zero results.

In a few language pairs, text that is formatted as a GUID (hex characters separated by dashes) sometimes return changed. This is particularly the case for language pairs that have different number formatting, like Chinese and some Indian languages. We have added GUID detection on our side, but it would be nice if the API didn't make it necessary.

Ich vermisse in der Übersetzung von Deutsch zu Französisch die Möglichkeit, eine generelle Einstellung auf die Anrede Du beziehungsweise Sie vornehmen zu können. Der Text wird fast immer in der Sie-Form übersetzt, so dass es keine korrekte Anrede ist.

Hi,

There a good number of Kannada Language content users and i'd love to see the addition of Kannada Language (kn) in the translation service.

Do you have the program's ocx application? for vb6, or visual net.

HELP! I am having trouble with the latest update for Microsoft 365 Office ProPLUS. Before the update, I was able to use Translator for live captioning with no issues. However, ever since the school district update to ProPLUS, I can't see the Subtitle icon under Slide Show ribbon. Please help!

According to https://docs.microsoft.com/en-us/azure/cognitive-services/translator/reference/v3-0-translate#obtain-alignment-information, the user 'will not receive alignment if the sentence is a canned translation. Example of a canned translation is 'This is a test', 'I love you' and other high frequency sentences.'

Could you please be consistent and provide a 'canned alignment' for canned sentences?

Always getting alignment information is quite critical for our integration of the Translator Text API.

Thanks.